Professional Documents
Culture Documents
www.mn-net.com
www.mn-net.com
www.mn-net.com
What does it take to solve
one of the biggest challenges
in chromatography?
waters.com/PREMIER
316 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
MANUSCRIPTS: For manuscript preparation guidelines, see chromatographyonline. 485F US Highway One South, Digital Production Manager
com/lcgc-author-guidelines, or call The Editor, (732) 596-0276. LCGC welcomes unso- Suite 210 Sabina Advani
licited articles, manuscripts, photographs, illustrations, and other materials but can- Iselin, NJ 08830 SAdvani@mjhlifesciences.com
not be held responsible for their safekeeping or return. Every precaution is taken to (732) 596-0276
Fax: (732) 647-1235 Managing Editor, Special Projects
ensure accuracy, but LCGC cannot accept responsibility for the accuracy of informa- Kaylynn Chiarello-Ebner
tion supplied herein or for any opinion expressed. KEbner@mjhlifesciences.com
PUBLISHING/SALES
SUBSCRIPTIONS: For subscription and circulation information: LCGC, P.O. Box
Senior Vice President,
457, Cranbury, NJ 08512-0457, or email mmhinfo@mmhgroup.com. Delivery of MARKETING/OPERATIONS
Industry Sciences
LCGC outside the United States is 14 days after printing. (LCGC Europe and Michael J. Tessalone Marketing Manager
LCGC Asia Pacific are available free of charge to users and specifiers of chro- MTessalone@mjhlifesciences.com Brianne Pangaro
matographic equipment in Western Europe and Asia and Australia, respectively.) BPangaro@mjhlifesciences.com
Associate Publisher
Edward Fantuzzi C.A.S.T. Data and List Information
CHANGE OF ADDRESS: Send change of address to LCGC, P.O. Box 457, EFantuzzi@mjhlifesciences.com Melissa Stillwell
Cranbury, NJ 08512-0457; alternately, send change via e-mail to MStillwell@mmhgroup.com
fulfill@mjhassoc.com or go to the following URL: Sales Manager
http://mmhpubs.mmhgroup.com/Welcome.aspx?pubid=LCGC Brianne Molnar Reprints
BMolnar@mjhlifesciences.com Alexandra Rockenstein
Allow four to six weeks for change. PUBLICATIONS MAIL AGREEMENT
ARockenstein@mjhlifesciences.com
No. 40612608. Return all undeliverable Canadian addresses Senior Director, Digital Media
to: IMEX Global Solutions, P.O. Box 25542, London, ON, N6C Michael Kushner Audience Development Manager
6B2, CANADA. Canadian GST number: R-124213133RT001. MKushner@mjhlifesciences.com Jessica Stariha
JStariha@mmhgroup.com
C.A.S.T. DATA AND LIST INFORMATION: Contact Melissa Stillwell,
EDITORIAL
tel. (218) 740-6831, e-mail MStillwell@mmhgroup.com.
CORPORATE
REPRINTS: Contact Michael J. Tessalone, Editorial Director
e-mail: MTessalone@mjhlifesciences.com Laura Bush Chairman & Founder
INTERNATIONAL LICENSING: Contact Kim Scaffidi, LBush@mjhlifesciences.com Mike Hennessy Sr
e-mail: kscaffidi@mjhassoc.com
Managing Editor Vice Chairman
John Chasse Jack Lepping
CUSTOMER INQUIRIES: Customer inquiries can be forwarded directly to MJH JChasse@mjhlifesciences.com
Life Sciences, Attn: Subscriptions, 2 Clarke Drive, Suite 100, Cranbury, NJ President & CEO
08512; e-mail: mmhinfo@mmhgroup.com Senior Technical Editor Mike Hennessy Jr
Jerome Workman
JWorkman@mjhlifesciences.com Chief Financial Officer
Neil Glasser, CPA/CFE
ycled P
Rec a Associate Editor
%
Cindy Delonas Executive Vice President,
pe
50
r
10-20%
CDelonas@mjhlifesciences.com
Corporate Development
rW
Po
s t C s u me
on Creative Director, Publishing Joe Petroziello
Melissa Feinen
mfeinen@mdmag.com Executive Vice President,
© 2020 MultiMedia Pharma Sciences, LLC. All rights reserved. No part of this publi- Operations
Senior Art Director Tom Tolvé
cation may be reproduced or transmitted in any form or by any means, electronic or
Gwendolyn Salas
mechanical including by photocopy, recording, or information storage and retrieval
gsalas@mjhlifesciences.com Senior Vice President, Content
without permission in writing from the publisher. Authorization to photocopy items
Silas Inman
for internal/educational or personal use, or the internal/educational or personal Graphic Designer
use of specific clients is granted by MultiMedia Pharma Sciences, LLC. for libraries Courtney Soden Senior Vice President, I.T.
and other users registered with the Copyright Clearance Center, 222 Rosewood Dr., csoden@mjhlifesciences.com & Enterprise Systems
Danvers, MA 01923, (978) 750-8400, fax (978) 646-8700, or visit http://www.copyright. John Moricone
com online. For uses beyond those listed above, please direct your written request
to Permission Dept. email: ARockenstein@mjhlifesciences.com Senior Vice President, Audience
CONTENT MARKETING
Generation & Product Fulfillment
MultiMedia Pharma Sciences, LLC. provides certain customer contact data (such as Joy Puzzo
Custom Content Writer
customer’s name, addresses, phone numbers, and e-mail addresses) to third parties
Allissa Marrapodi
who wish to promote relevant products, services, and other opportunities that may Vice President, Human
AMarrapodi@mjhlifesciences.com
be of interest to you. If you do not want MultiMedia Pharma Sciences, LLC. to make Resources & Administration
your contact information available to third parties for marketing purposes, simply Webcast Operations Manager Shari Lundenberg
email mmhinfo@mmhgroup.com and a customer service representative will assist Kristen Moore
you in removing your name from MultiMedia Pharma Sciences, LLC. lists.
Vice President, Business
KMoore@mjhlifesciences.com
Intelligence
LCGC North America does not verify any claims or other information a ppearing in Project Manager Chris Hennessy
any of the advertisements contained in the publication, and cannot take responsi- Vania Oliveira
bility for any losses or other damages incurred by readers in reliance of such content. Executive Creative Director,
VOliveira@mmhgroup.com
Creative Services
To subscribe, email mmhinfo@mmhgroup.com. Jeff Brown
Quantification of Adeno-
Associated Virus Aggregation
The Field-Flow Fractionation Platform
Adeno-associated viruses (AAVs) are increasingly used for gene therapy due
to their versatility and safety. One of the biggest concerns for manufacturing
a uniform AAV suspension is the presence of viral aggregates, which
can create problems with transduction efficiency, biodistribution, and
immunogenicity. These large AAV aggregates are challenging to separate
and characterize by traditional column-based chromatography techniques
such as size exclusion chromatography (SEC).
NovaFFF Software
MALS DLS UV RI ICP-MS
CONTENTS
COLUMNS
320 LC TROUBLESHOOTING
Recovering from a COVID-19 Shutdown:
Tips and Tricks for Starting Up, Part I
June 2020
Size in GC After a Pandemic Shutdown
371 FUNDAMENTALS
Troubleshooting Gas Chromatography:
Reduced Peak Size (Loss of Sensitivity)
Tony Taylor
There are many potential causes of reduced peak size in gas
chromatography (GC), and an inexperienced GC user may not
know where to begin the troubleshooting process. Here, we
review potential causes for reduced peak size in GC systems.
FEATURE ARTICLE
LCGC North America (ISSN 1527-5949 print) (ISSN 1939-1889 digital) is published monthly by MultiMedia Healthcare, LLC., 2 Clarke Drive, Suite 100, Cranbury, NJ 08512, and is
distributed free of charge to users and specifiers of chromatographic equipment in the United States and Canada. LCGC is available on a paid subscription basis to nonquali-
fied readers in the United States and its possessions at the rate of: 1 year (13 issues), $83.95; 2 years (26 issues), $151.20; in Canada and Mexico: 1 year (13 issues), $107.10; 2 years
(26 issues), $168.53; in all other countries: 1 year (13 issues), $157.50; 2 years (26 issues), $281.40. Periodicals postage paid at Trenton, NJ 08650 and additional mailing offices.
POSTMASTER: Please send address changes to LCGC, P.O. Box 457, Cranbury NJ 08512-0457. PUBLICATIONS MAIL AGREEMENT NO. 40612608, Return Undeliverable Canadian
Addresses to: IMEX Global Solutions, P. O. Box 25542, London, ON N6C 6B2, CANADA Canadian GST number: R-124213133RT001. Printed in the USA.
WWW.CHROMATOGRAPHYONLINE.COM JUNE 2020 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 319
WinGPC UniChrom
GPC/SEC, IPC and 2D of
(Bio)Polymers and Proteins
Macromolecular Chromatography
Data System
Ensure Compliance and
Data Integrity Data Evaluation PERFECT
Instrument control for all major SEPARATION SOLUTIONS
LCs and specialty detectors MultiWorkstation
Request a Live Demo at your Desk:
Data Evaluation, Client/Server info@pss-polymer.com
or MultiWorkstation Client/Server www.pss-polymer.com
320 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
LC TROUBLE SHOOTING
process of restarting their laboratories. I’ve tative measures can save a lot of time fer first with water, and then the
invited Tony Taylor to join me in sharing some and money later on when you are ready organic solvent.
of the advice and procedures he has devel- to start collecting useful data again.
oped over the years. Although we’ve never Remember that the column is the heart Solvents and Buffers
worked through pandemic-related laboratory of any LC system, and arguably the most Figure 1 shows a picture of a buffer bot-
shutdowns and startups, some of our expe- sensitive to potential shutdown-related tle I observed in an LC laboratory about
riences with troubleshooting liquid chroma- problems. Be sure to disconnect the a year ago. The solution was dilute (1
tography (LC) problems during normal times column from the flow path before doing mM) phosphoric acid in water. Even
should prove very useful under these strange anything else with the system, and only though this is fairly acidic (pH ~2.7),
circumstances as well. reconnect it to the flow path after you there was a very healthy microbial com-
Dwight Stoll are sure that the rest of the system (that munity growing inside (affectionately
Bio Fermented Alcohol Detection
Utilizing Isocratic Cation Exchange
HPLC with Hamilton’s HC-75 Column
Concern over detrimental greenhouse gases produced from the and lead to less carbon monoxide side products.2 Production of ethanol
combustion of fossil fuels has been growing for the past few decades. from biological sources produce side products, specifically propanol
Identification of new and more diverse methods of energy use has been and methanol amongst other ketones, fatty acids, and esters. Due to
investigated around the world. One such avenue is combustible alcohols azeotropic formation of ethanol with water it is difficult to distill pure
produced from biomasses like sugar beet or sugar cane. To reduce 200 proof ethanol products. However, HPLC analysis provides a good
dependence on fossil fuels, a concerted effort has been made to develop platform for ethanol determination after biomass fermentation.3 To aid
processes where biomass can be fermented biologically by bacteria or the detection of ethanol purity, Hamilton has developed a method utilizing
yeast to produce combustible alcohols. Combustible alcohols are desired the HC-75 (H+ Form) HPLC column, which provides good separation
over other technologies due to the convenience of switching from petrol of the desired analytes. With ~8% crosslinking of the PS-DVB backbone
based fuels. Ethanol is already added to most gasolines to increase the resulting porous gel stationary phase matrix allows for enhanced
octane efficiency. With anywhere from 5–25% ethanol being added to interactions between stationary phase and analyte. Detection was
petrol products in most of the world and up to 100% ethanol fuel in Brazil.1 accomplished with refractive index, but other methods of detection
Additionally, the combustion of these alcohols tend to be more efficient can be applied (i.e. UV, or mass spectrometry).
Column Information
Packing Material HC-75 (H + Form), 9 µm 3
P/N 79544
Chromatographic Conditions
2
Gradient Isocratic
Temperature 80°C
1
Injection Volume 10 µL
1) E. Gnansounou, A. Dauriat, Ethanol fuel from biomass: A review. J. Sci. & Indust. Res.
(64) 2005, 809–21.
2) S. Onuki, J. A. Koziel, W. S. Jenks, L. Cai, D. Grewell, J.H. van Leeuwen, Taking ethanol
quality beyond fuel grade: A review. J. inst. Brew. (122) 2016, 588–98.
3) S.Y. Lee, R. Sankaran, K.W. Chew, et al. Waste to bioenergy: a review on the recent
conversion technologies. BMC Energy. 1:4, 2019.
©2020 Hamilton Company. All rights reserved.
All other trademarks are owned and/or registered by Hamilton Company in the U.S. and/or other countries.
Author: Adam L. Moore, PhD, Hamilton Company Lit. No. L80105 — 05/2020
Web: www.hamiltoncompany.com Hamilton Americas & Pacific Rim Hamilton Europe, Asia & Africa
Hamilton Company Inc. Hamilton Central Europe S.R.L.
USA: 800-648-5950 4970 Energy Way
Reno, Nevada 89502 USA
str. Hamilton no. 2-4
307210 Giarmata, Romania
Europe: +40-356-635-055 Tel: +1-775-858-3000 Tel: +40-356-635-055
Fax: +1-775-856-7259 Fax: +40-356-635-060
To find a representative in your area, please visit hamiltoncompany.com/contacts. sales@hamiltoncompany.com contact.lab.ro@hamilton-ce.com
322 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
to simply not use that solvent bottle any case, if the solvent inlet filter is
again for that instrument, because if porous—whether sintered glass or stain-
thoroughly cleaning the bottle requires less steel—just throw it away, and replace
detergent or other cleaning agents, the it with a new one. These porous filters are
potential for problems associated with perfect environments to support micro-
those cleaning agents is greater than bial growth, and getting them truly clean
the benefit of having a clean bottle to after they have been contaminated is
work with. For example, some deter- extremely difficult. Inlet filters that are
gents can be difficult to remove from based on screens or mesh are more eas-
glassware, to the point where they are ily cleaned. Remove the filter from the
not detectable by MS, and thus what solvent line, remove the visible debris
seems like a simple bottle cleaning step with a brush, and then soak in detergent,
can turn into a background contaminant followed by several rinses with water,
peak in the MS that is detectable for and finally organic solvent. Sonicating
months. When using MS detection, we sintered glass filters should be avoided,
feel the best solution to a badly con- but sonicating plastic or metal ones is
taminated solvent bottle is a brand new okay. As with the bottle, rinse the filter a
FIGURE 1: Image of an LC solvent bottle solvent bottle. In other words, just get few times with the solvent you intend to
with visible (cloudy) microbial growth. rid of the contaminated bottle without use it with before putting it in the clean
bothering to try to clean it. bottle. For the solvent line itself, it is best
known as “floaties”). The good news On the other hand, if the contami- to remove the line from the pump and
in this particular case is that the scien- nated bottle is on an instrument using attach a syringe (10 mL or so will do) so
tists were able to recover their pumping a less sensitive detector (for example, that you can pull large quantities of rins-
system without any long-term effects; ultraviolet-visible [UV-vis] spectros- ing solvent through the line. For older
however, the column they were using copy or an evaporative light scatter- systems that have degasser modules
at the time suffered a performance loss, ing detector [ELSD]), then the bottle separate from the pump itself, discon-
and was never quite the same after- usually can be cleaned without major nect the solvent line at the pump so that
ward. We expect that this experience impact on the rest of the system. After you can pull cleaning solvents through
is going to be repeated in hundreds of discarding the contaminated solvent, the degasser also. Pulling detergent
laboratories as a result of the COVID- wash thoroughly with a detergent solution through the degasser should
related laboratory shutdowns. Many of designed for use with glassware, rinse be avoided, because this may persist in
the aqueous buffer solutions used in four times with HPLC-grade water, and the tubing for a long time. We find that
LC (and even high-performance liquid finally with HPLC- grade organic sol- using warm water (up to 60 °C) is quite
chromatography [HPLC] grade water, vent. We find that working through a helpful for flushing the internal tubing
if given enough time and exposure to series of organic solvents of increasing of the degasser. For newer pumps with
laboratory dust) are environments quite dipolarity—such as cyclohexane, aceto- integrated degassers, it should be suf-
favorable to microbes, particularly those nitrile, and methanol (in that order)—is ficient to disconnect the solvent line at
in the middle of the pH range. Simply effective for removing contaminants of the inlet to the degasser on the pump
stated, many scientists are going to varying degrees of dipolarity. Before unit. Using the same steps as with the
return to laboratories where they will be use again on the instrument, be sure to bottle and filter, pull each liquid through
able to see visible microbial growth in rinse with whatever solvent you plan to the solvent line as a means of rinsing the
the solvent bottles of their LCs. put in the bottle. inside of the line. For older systems with
So…what to do? Now, if you come back to your labo- degassers that have large internal vol-
Occasionally we have problems in our ratory and find microbes in your solvent umes, something on the order of 50 mL
own laboratories with solvent bottles bottles, then the solvent line feeding the of each liquid will be enough to give the
that become contaminated somehow, pump is also contaminated, in addition solvent line and degasser a good flush.
whether resulting from a bad lot of sol- to the solvent bottle itself. Before put- With newer systems where you only need
vent from a vendor, or an instrument ting this solvent line back into the solvent to flush the line external to the pump; 10
operator pouring something into the bottle you have just cleaned, consider mL or so should be good enough.
bottle that does not belong in there. what exactly you will do with the solvent
Whenever this happens, our approach line. Again, if this is an LC–MS instrument, Dirty and Dried-Out Pumps
depends on whether or not the instru- you should seriously consider replacing Before attempting to move any liquid
ment is using mass spectrometric (MS) the entire solvent line and solvent inlet through the pump itself, it is best to dis-
detection. If it does, our preference is filter that goes in the solvent bottle. In connect the high pressure outlet capil-
WWW.CHROMATOGRAPHYONLINE.COM JUNE 2020 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 323
lary from the pump and connect a waste valve. These can also suffer from partial care to be sure that, if the rest of the sys-
line running to a waste container to col- or total obstruction, leading to inconsis- tem contains a buffer solution, the first
lect whatever goes through the pump tent flow (if on the inlet side), or higher solvent pumped from the clean pump
until you are confident that the pump is than expected pressure (if on the outlet to the rest of the system is water, so that
delivering clean, debris-free solvent. If it side). If you know you are dealing with precipitation of salts can be avoided.
is obvious that one or more solvent lines a situation where it is likely these filters
has been contaminated with microbes, or screens have been fouled due to pro- Component Blockages Due
it will be best to flush the pump thor- longed inactivity of the pump, microbial to Precipitation and Debris
oughly with clean solvent before con- growth, or other debris, it will be a good Some users will return to their labora-
necting the pump to the rest of the idea to simply replace these filters or tories to find that there is a significant
system. Our approach to this step is to screens at the beginning of your startup obstruction somewhere in their LC sys-
put all of the solvent lines associated process. If you want to try cleaning them tem that is preventing flow at reason-
with the pump (typically two for a binary before replacing, remove and sonicate able pressures. This obstruction could
pump or four for a quaternary pump) in warm water for 10 min, followed by result from debris associated with con-
into a single bottle of HPLC-grade IPA for 10 minutes. If problems persist taminated solvents, pump or valve seal
water, and run 30 mL of water through that are attributed to these parts, just material if salt precipitation occurs in
each channel of the pump. Next, put all replace them entirely. those components, or precipitation of
of the solvent lines into a single bottle If the pump has dried out entirely salts inside of some other components
of isopropyl alcohol (IPA), and run 30 mL because it was left on with an empty such as a connecting capillary. The first
of it through each channel. Finally, move solvent bottle, it can be difficult to step in resolving such obstructions is
each solvent line back to its respective remove air bubbles from pump heads to clean the pump as discussed above,
solvent bottle (for example, one in water, during startup. Air bubbles can also and make sure it will pump water with
one in buffer, and one in acetonitrile), accumulate in pump heads if the sys- a waste tube connected directly to
and flush each channel with 30 mL of the tem is not used for an extended period the high pressure outlet of the pump.
solvent intended for use in the method. of time. In our experience, the best way Then, systematically add one connec-
It is important to recognize that there to resolvate a completely dry pump or tion or component at a time, working
are some filters in modern LC systems remove air bubbles from the pump your way from the pump outlet toward
that might be easy to overlook. It is heads is to purge thoroughly with IPA, the detector. In most systems, the next
always helpful to consult the technical because it wets most materials used in component after the pump will be the
manual for your LC system to see where LC pumps well. autosampler, with a fairly long connect-
these might be. For example, some Once the pump has been thoroughly ing capillary between them. Disconnect
pumps have porous frits or screens on purged with clean solvent and it is deliv- this capillary from the sampler, con-
the upstream side of the inlet check ering the expected flow from each chan- nect it to the pump, and see if you can
valve. Other pumps have porous frits on nel at low pressure, it can be connected pump water through this capillary at
the downstream side of the outlet check to the rest of the system. Again, take reasonable pressure; what pressure is
Unbeatable accuracy
again and again
The precise and time-pressured nature of your work
mean you need products you can rely on, every time.
That’s why every product across the comprehensive
Supelco® portfolio is meticulously tested against the
tightest specifications to ensure the highest industry
standards and compliance with applicable regulations.
So for unbeatable accuracy every time, choose
Supelco® analytical products.
MilliporeSigma, the vibrant M and Supelco are trademarks of Merck KGaA, Darmstadt,
Germany or its affiliates. All other trademarks are the property of their respective owners.
Detailed information on trademarks is available via publicly accessible resources.
© 2020 Merck KGaA, Darmstadt, Germany and/or its affiliates. All Rights Reserved.
The Life Science Business of Merck KGaA, Darmstadt, Germany operates as
MilliporeSigma in the US and Canada.
324 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
reasonable depends on the dimen- mission of light through the flow cell.
THEIRS IS sions of the capillary. A benchmark to Most manufacturers recommend flush-
FAMILIAR. remember is that the pressure drop ing the flow cell with warm water (up
PER SPEC TI V E S
IN MO DERN HPLC
dard operating procedures (SOPs) to sup- In this paper, we use solid oral dos- into a commercial drug product, and sub-
plement regulatory guidelines to provide age forms, such as tablets or capsules, as mitted for regulatory approval.
more specific details, to ensure that stabil- examples. Four types of testing for orally In most pharmaceutical companies, the
ity studies are appropriate for their specific available products, listed in Table II, are nomination of an NCE to the status of a
product types. Nevertheless, stability pro- discussed. Biological products are not drug development candidate triggers the
grams and their practices can vary widely, covered here. formation of a multidisciplinary technical
particularly between large and small Other dosage forms such as parenterals development team. This team is responsi-
pharmaceutical companies, often due to also have similar types of tests, including ble for taking the drug candidate into clini-
the depth of knowledge and available identity, chemical testing (assay and impu- cal trials in humans, and eventually to drug
resources. The primary aim of this paper is rities), physical testing (pH, clarity, particu- approval, production, and commercializa-
to increase understanding of the science, lates), and sterility tests as appropriate. tion. The team is responsible for activities
326 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
• finding the optimum solid-state form the quality of CTMs. This method is typi- Good Manufacturing Practices [GMP]) sup-
(salt, crystallinity) or polymorph of the cally a “composite” analytical procedure porting the establishment of expiration
API (analytical staff) that determines both the potency of the dates for regulatory filings (5,6). Details on
• conducting stability studies to support API and chemical impurities of the DS. Ide- the development and validation of the sta-
clinical development and registration ally, it can also be used for the assessment bility-indicating methods will be covered
(analytical and QC staff) of DP samples. Once a feasible analytical in the second and third installments of the
• designing and developing processes for procedure is established, forced degra- series. Forced degradation and stability
formulations of CTMs and the final com- dation studies are performed to evaluate studies are discussed in later sections here.
mercial DP (pharmaceutics staff) the specificity of the analytical procedure. A better understanding of stability
• setting acceptance criteria for CQAs This method is used in the release testing studies and their regulatory requirements
(specifications) to monitor clinical quality of the early clinical batches and testing of is essential for the analytical chemist for
and establishing commercial specifica- samples from initial stability studies to sup- several reasons (6,7). First, these stud-
tions (QA and regulatory staff) port clinical trials (3–5). ies require many different tests (shown
• manufacturing CTMs (CMC team and out- The stability-indicating assay and in Table II) on multiple batches stored at
sourcing and project management staff) impurities method is typically a gradient different storage conditions and pulled at
• assembling the CMC submission pack- reversed-phase HPLC method with UV numerous time intervals. Second, the pro-
age for regulatory filings (CMC team detection that can separate the API and cess development or formulation scientists
and regulatory staff). all the known impurities and degradation may request many experimental batches
In this series of white papers on stability products (5). This method is validated to of DS or formulations to be placed on sta-
testing, we focus on the tasks performed ensure adequate analytical performance bility, leading to redundant testing, bring-
by the analytical development and QC (specificity, linearity at 100% target concen- ing little added value if not managed with
scientists within the CMC team. The first tration and expected impurities concentra- a science- and risk-based approach. There-
task for the analytical chemist assigned to tions, accuracy, precision, sensitivity, and fore, a better understanding of the design,
the CMC team is to develop a reasonable solution stability) for “formal stability stud- intent, and best practices of stability stud-
stability-indicating method for monitoring ies” (defined as studies conducted under ies and a proper interpretation of associ-
Unrivalled separation power Micromachined reproducibility Flow rate flexibility Plug-and-play connectivity
1 - 15 µl/min
such as an Investigational New Drug (IND), ment process, according to ICH Q1A (R2) Forced degradation studies are per-
New Drug Application (NDA), or Abbrevi- and ICH Q2 (R1) (8,16). Although regula- formed to investigate the major degrada-
ated New Drug Application (ANDA) filing (1). tions mandate that these studies must be tive pathways of the DS and DP and support
The stability data packages within the CMC performed, the guidance does not provide the initial development of the DS stability-
section are typically the most extensive non- directions on procedures or conditions to indicating methods (3,5,12). High-stress
clinical sections in these submissions. They use. The actual protocols employed appear conditions are employed to subject DS and
include tabulated data, graphs, narratives to to vary widely in different organizations. An DP to conditions more severe than acceler-
summarize the stability profile of the pack- overview of the chemistry fundamentals of ated conditions to serve several purposes:
aged product to justify specifications, and drug forced degradation studies and best • to provide insight into the degradation
proposed expiry for the product. practices can be found in books, articles, pathways of DS and DP and their mech-
and other resources (17–20). anisms, facilitating the development
Harmonization of Regulations: ICH
Quality Guidelines for Stability Studies
TABLE V: An elaborate forced degradation protocol used in an automated laboratory Thermal Degradation
The effect of temperature is studied by expos-
Stressing Condition Temperature (°C) Stressing Time (days)
ing the drug substance to high temperatures
Water 70 0, 0.25, 0.5, 0.75, 1, 1.5, 2, 3, 5, 7 in increments of 10 oC from 50 to 80 oC. Ther-
0.1 N HCl 70 0, 0.25, 0.5, 0.75, 1, 1.5, 2, 3, 5, 7 mal degradation is the primary degradation
0.1 N NaOH 70 0, 0.25, 0.5, 0.75, 1, 1.5, 2, 3, 5, 7 pathway of the DS in the solid state.
0.3% H2O2 Ambient 0, 0.25, 0.5, 0.75, 1, 1.5, 2, 3, 5, 7, 10, 14
Solution State Acid/Base Degradation
Cool White Fluores-
Ambient 0, 0.25, 0.5, 0.75, 1, 1.5, 2, 3, 5, 7, 10, 14 Solution-state degradation mechanisms may
cent Light (solution)
Cool White Fluores- be different, and should also be explored.
Ambient 0, 1, 2, 3, 5, 7, 9, 10, 11, 12, 13, 14
cent Light (solid) If the drug substance is not soluble under
Suntest—UV and aqueous conditions, then a small amount
Ambient 8 h (1xICH), 16 h (2xICH)
Visible Light (solid) (up to 10%) of organic solvent (acetonitrile
PH 2 Buffer 70 0, 1, 3, 5, 7 or methanol) can be used to solubilize the
PH 4 Buffer 70 0, 1, 3, 5, 7 DS, and the acid or base can then be added
PH 6 Buffer 70 0, 1, 3, 5, 7 to stress the samples. Attention should
be given to the functional groups pres-
PH 8 Buffer 70 0, 1, 3, 5, 7
ent in the drug molecule when selecting
PH 10 Buffer 70 0, 1, 3, 5, 7
a co-solvent (20).
Week 1: 60 It is not advisable to use a high concen-
In solid form at
Week 2: 70 0, 1, 3, 5, 7, 8, 10, 12, 14, 15, 17, 19, 21
ambient humidity
Week 3: 80
tration of acid or base, because doing so is
neither practical nor necessary. A few papers
Week 1: 60
In solid form at 75%RH Week 2: 70 0, 1, 3, 5, 7, 8, 10, 12, 14, 15, 17, 19, 21 have suggested a series of extensive steps
Week 3: 80 of pH changing from pH 2 to pH 10 buffer. It
may not be necessary as a routine practice,
unless for a specific DS or as part of a stress
Design of Experiments (DoE) to establish a
TABLE VI: Typical forced degradation studies for drug product of a solid-dosage form
stability model (21).
Forced Degradation Study Suggested Conditions
Heat Expose drug products at 50 ºC for up to 1 month Photostability
Expose drug products to 40 ºC and Photostability studies are essential, and
Humidity
75% RH and 25 ºC and 90% RH must be performed to generate primary
Photostability
Expose drug product in a petri dish without light-induced degradation products by
cover at twice or three times the ICH Q1B exposure level subjecting one representative batch of the
DS to the light exposure listed in ICH Q1B.
The standard light exposure is 1.2 million
of stable formulations and selection of in both solid and solution states. Table IV lux hours of visible light and 200 lux hours
suitable packaging lists an example set of stress conditions of near UV, with storage conditions con-
• to obtain samples to verify method for DS. An important goal of forced deg- trolled at room temperature.
specificity and structure elucidation of radation studies is to generate a potential These studies can be repeated when
significant degradation products level of degradation products that may there is a change of DS samples, such as a
• to allow the differentiation of impurities form during manufacturing and on stabil- new synthetic route, with different crystal-
and degradation products from the DS, ity storage. Typically, forced degradation linity form, different supplier, or a change
excipients, and other interference studies are conducted until a 5–20% loss of DP formulation.
• to facilitate the rapid establishment of of API is observed (18,19). This range is set During the method development phase,
CQAs for selection or elimination of to produce a reasonable amount of deg- forced degradation samples are typically
selected testing. radation products to facilitate method evaluated using a preliminary mass spec-
• to rapidly identify any excipient incom- development and to avoid secondary trometry (MS) compatible HPLC method
patibility issues with the DS degradation products from an unstable with a photodiode array detector (PDA)
• to generate potential data to support degradant (which are not observed under to collect information on peak area under
the justification of specifications. actual stability conditions). It should be the curve. The MS and PDA data are used
Forced degradation studies are con- noted that the conditions should be for peak tracking, peak purity assessment,
ducted under high temperature, humidity, selected based on the physicochemical identification of degradation products,
acid/base, oxidative, and light conditions properties of the individual DS and DP. impurities, and interferences (3,5,19).
WWW.CHROMATOGRAPHYONLINE.COM JUNE 2020 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 331
Table V shows an example of a more extensive, elaborate forced these chambers be controlled within ±2 oC and ±5% RH for
degradation protocol used by a pharmaceutical laboratory to col- chambers operated at 25 °C/60% RH, 30 °C/65% RH, and 40
lect data for most of their NCEs using a four-temperature thermal °C/75% RH. The refrigerated condition is controlled at 5 oC ± 3
stressing condition matrix (ambient, 60 oC, 70 oC, 80 oC) for acid, oC (2 to 8 oC) with ambient humidity (monitored but not con-
base, oxidative, light, humidity, buffered solution, and prolonged trolled). These chambers are equipped with sensors for the
stressed duration (up to 21 days). This sample-intensive protocol continuous monitoring of temperature and relative humidity.
was made feasible by this centralized and automated laboratory These data are captured electronically or by chart recorders
specializing in supporting forced degradation studies for the with backup power. Excursions will be noted with an alarm
development site. The laboratory utilized automated sample stor- system. When the chambers are out of tolerance (see above),
age and retrieval system called “Powderium” (a powder dispens- investigations must be performed to identify the cause of the
ing workstation), robotics, programmable liquid handling, and issue and determine the impact of the stored samples (7).
refrigerated storage of quenched samples (18,19). The stability chambers are fully qualified and maintained, and
Table VI provides a summary of forced degradation studies per- their access is limited to authorized personnel to ensure that sam-
formed on solid dosage forms. Note that acid, base, and oxidative ple removal is tracked and controlled. Maintenance and service
exposures are not used, because these conditions are not repre- records must be available for inspection during audits. Downtime
sentative of realistic storage or exposure situations. The length of must be limited, and a backup plan must be developed to avoid
exposure should be evaluated carefully, depending on the dosage issues that can affect sample storage and data integrity. An annual
forms or the formulation. For example, capsule formulations or chamber inventory program is also necessary to track the sample
film-coated tablets may not be able to withstand a temperature storage and disposition at any time.
higher than 60 oC for an extended length of time.
Similar to the DS, a goal of 5% to 20% loss of active ingredient Stability Study Strategies
is recommended. If DS or DP is stable and the desired degrada- to Expedite “First-in-Human” Clinical Trials
tion level cannot be reached under high-stress conditions, stud- Minimizing “Time to Market” is an essential strategy for most
ies should be terminated. Overstressing a sample may lead to pharmaceutical companies (1). Given that by regulatory agencies
secondary degradation products that are not present in formal expect to see three months of stability data in IND filings to enable
stability studies. In contrast, understressing may lead to insufficient
degradation products for method development or identification
(12). For each condition of the forced degradation study, the chro-
matogram is compared with that of a control sample to document
the mass balance of the API (% loss of API) and % degradation
(normalized peak area %). For all conditions, the peak shapes
and peak purity (by PDA and MS) of all components should be
assessed for coelution or presence of interferences (3–5).
Accelerated and stress studies are performed to induce quicker
degradation than what would be observed in recommended stor-
age conditions. Results from these studies are used to estimate the
stability profile of the DS and DP at long term storage conditions,
establish the actual degradation pathways, and demonstrate the
intrinsic stability of the DS. It is recommended that these studies
be implemented under GMP conditions because the results of
these studies may be used as supporting information in regulatory
document packages. Full shelf-life studies are still needed to verify
the retest date of the DS and expiration date of the DP.
While discussions on forced degradation studies for NCEs in
NDA applications are generally plentiful, those for ANDA are often
insufficient and may result in regulatory deficiencies. For instance,
the impurity profiles may not be completely derived or sufficiently
discussed (20). In addition, the labeling for generic drug products
should be concordant with that of the reference listed drug (RLD)
and with a compendial monograph, as applicable (20).
Stability Storage
Stability samples are kept in a controlled temperature in humid-
ity stability chambers. The ICH Q1A (R2) guideline requires that
332 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
closure systems. This approach is also found acceptable by many size) encompass and represent the stability behaviors of products of
regulatory agencies (23–25). the intermediate levels, eliminating the need for testing at the inter-
Improved modeling tools have enabled stability predictions mediate levels. A bracketing schedule can be applied to multiple
with accelerated timeframes compared to those of the traditional strengths of identical or closely related formulations.
extrapolation approach. Besides providing a more accurate way Table VIII shows an example in which the 100 mL container
to justify an expiry assignment, these models are used to increase size is bracketed by the 50 mL and 250 mL container sizes; thus,
the understanding of the critical factors that influence the quality stability samples stored in the 100 mL containers are not tested.
of the DS and DP, as demonstrated in empirical science and risk- Similarly, the 100 mg and 250 mg tablet strengths are bracketed
based approaches detailed in ICH Q8–Q11. These approaches by the 50 mg and 500 mg strengths. Assuming that the stability
are often utilized by large pharmaceutical companies with bet- profiles of these different strength formulations are the same
ter modeling resources and expertise. A regulatory template and the compositions of these tablets are similar, regardless of
has been shared to standardize on critical elements to use risk- strength, then the testing of 100 mg and 250 mg tablets may
based predictive stability data for setting up shelf life to support not be needed. ICH Q1A (R2) requires that three batches (A, B,
clinical development. (26) C) be made available for submission. According to the guide-
lines, a pharmaceutical company could reduce its testing from
Trending Analysis potentially 36 configurations to only 12 configurations, resulting
The purpose of the stability program is to establish a retest period in significant savings of resources and time (6,7).
for DS or shelf life for DP that will apply to all future commercial Table IX lists some factors to which a bracketing approach can
batches. Stability data of CQAs of individual batches should potentially apply, assuming that the container–closure system
remain within specification throughout the assigned retest period and the headspace of the containers do not have any impact.
of the DS or shelf life of DP. Therefore, the trending of all stability The disadvantage of the bracketing concept is that when one of
data is very important. Most degradation trends of API are linear; the results is out-of-specification; all bracketed configurations
however, the degradation pathway could be a linear, quadratic, or data could then be at risk. Full testing would need to be acti-
cubic function. The typical ICH Q1E approach for the determina- vated mid-study, complicating the data set, and negating the
tion of shelf life through the analysis of data based on a quantita- benefit of any testing reduction. In addition, all bracketed fac-
tive attribute that is expected to change over time is to determine
the period of time at which the 95% one-sided (or two-sided) con-
fidence limit for the mean curve intersects the acceptance crite-
rion (27,28). The annual commitment lots should also be evaluated
against the product trend to assure the consistency of the stability
profile of the DP and verify if there is any unintended change that
may impact the DP.
Bracketing
Bracketing refers to a study design in which only the extreme vari-
ables, such as extreme strengths, container sizes, or container fills,
are tested. The plan assumes that the stability behaviors of prod-
ucts manufactured or packaged at these extreme levels (such as, for
example, highest vs. lowest strength, or largest vs. smallest package
334 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
statistical extrapolations and risk assess- TABLE XIII: Questions about the conclusions of a stability report
ment for the design (6,7).
Table X shows an example of a two-
thirds factorial matrixing design. In this ► What is the stability profile of the product?
► Are all lots within the stability trend of the product?
schedule, there are two strengths and ► Can differences be attributed to a particular factor such as packaging or strength?
three batches made of each strength. ► Do all tests meet specifications?
All presentations are tested at time zero, ► Is there any test closely monitored? Are there any tests to be performed
more often?
12 months, and the end of the study (36 ► Is there any package closely monitored? Is there any package to be tested
months) for the highest possible precision more often?
as these time points are critical to the DP ► Do all tests meet specifications?
► Do the data support the expiration date?
stability profile. For all other time points (3, ► Are there any stability trends?
6, 9, 18, and 24 months), only two-thirds ► Discuss any out-of-specification investigation.
of the presentations get tested at a rotat- ► Discuss any statistical evaluation of the data set.
► ould the stability database be satisfactory to the proposed or approved expiration
W
ing schedule. Therefore, all configurations dating?
are tested at different time points for the
same amount of times in total.
The matrixing approach requires that
all samples be placed in stability cham- lifecycle state (such as NDA, ANDA, or according to ICH Q1E or statements
bers, including at the timepoints that supplement NDA). The stability data- that no overall trends and little variabil-
are reduced. Therefore, when there is an base comprises laboratory data gen- ity are observed in the dataset. Table
instability issue, the testing schedule can erated and stored electronically or as XIII lists typical questions that should
be changed to test all configurations. If printed reports. These stability data are be addressed in the stability conclusion.
some data points are not available, there entered into stability reports, which are “Poolability” of different batches should
will be a minimum impact on the entire essential to communicate information be evaluated to determine the consis-
study of multiple presentations. internally for audit purposes, product tency of the representative stability data
Table XI lists some examples of factors investigations, justification of specifica- and trends with respect to the manufac-
where matrixing can be used to reduce tions, to support product development, turing process and analytical method,
testing. Matrixing also provides more or simply as a communication tool. All by comparing the intercepts and the
flexibility because different reduced detailed information in the report must slopes of all batches (6,7,24).
testing options can be performed for be verified for accuracy. For an NDA or
various tests depending on the test- ANDA submission, the stability report is Summary and Conclusions
ing confidence. Realistically, matrixing typically the most extensive non-clinical A robust and science-based stability
should be used only for label storage section to summarize the stability data program is required for the registration
conditions. The accelerated stability data of the product during its shelf life (6,28). and commercialization of any pharma-
may provide some insight into the sta- Stability data can be presented in a tab- ceutical product. Regulations in this
bility profile of the product at the label ular format, in a graphical representation, area are well established; however, the
storage conditions. Because matrixing is a narrative, or a combination of formats. application and interpretation of the
a statistical concept, evaluation of data Table XII is an example of a stability data regulations vary between companies
can be more involved, and assessment record. The data tables can be generated and regions. ICH harmonizes the main
of all the presentations as a complete manually in a document or printed out regions to help companies use a stan-
set is more complicated, which explains from a laboratory information manage- dardized approach.
why this approach is not used widely. ment system (LIMS) or stability manage- This article provides a high-level
These reduced testing approaches are ment software. All information and data summary of regulations and regula-
discussed in detail in many global guide- must be reviewed and verified. The con- tory expectations of stability programs.
lines; however, a lack of knowledge of clusion of the stability report discusses It reviews the goals and practices of
matrixing among regulatory reviewers whether the data indicate that the prod- forced degradation studies to generate
and internal quality and regulatory pro- uct continues to meet the quality specifi- sample data to confirm method speci-
fessionals may hinder the use of this cations or acceptance criteria established ficity. A strategy for conducting accel-
application to reduced stability testing. for the study and whether the data sup- erated stability studies on preliminary
port the proposed or approved expiration batches to expedite initial regulatory
Stability Reports and dating period (6,7). filing is described. Finally, predictive
Regulatory Submissions Stability reports should include a sta- stability software-based and statistical
Stability reports are required in all phases tistical evaluation of the currently avail- approaches such as trending, bracket-
of regulatory submissions regardless of able data and a discussion of the results ing, and matrixing are explained.
336 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
Acknowledgments (12) H. Bruemmer and N. Belikova, Pharm. (26) H. Williams, Pharm Technol. 42(8),
The authors express their gratitude to Technol. 41(9), 46–52 (2017). 42– 47 (2018).
the following colleagues for their time (13) 21 Code of Federal Regulations (CFR), (27) WHO Expert Committee on Specifica-
Part 211.194(a), Current Good Manufac- tions for Pharmaceutical Preparations
and efforts to provide timely reviews of turing Practice for Finished Pharmaceu- Fifty-second report, annex 10, Stabil-
the manuscript to improve content accu- tical Products (US Government Printing ity testing of ac tive pharmaceutical
racy and clarity: Jane Weitzel of Zinata, Office, Washington, DC, 2019). ingredients and finished pharmaceuti-
(14) United States Pharmacopeia Conven- cal products (World Health Organiza-
Jing Capucao of J&J, Mike Shifflet of tion, 2018).
tion, USP 37/NF 32, General Chapter
Johnson and Johnson Consumer Health USP <1225>, Validation of Compendial (28) International Council for Harmonisa-
Care, Adrijana Torbovska of Farmahem, Methods, (United States Pharmacopeial tion of Technical Requirements for
Alice Krumenaker of TW Metals, LLC; Convention, Rockville, Maryland, 2013). Pharmaceuticals for Human Use (ICH)
(15) United States Pharmacopeia Conven- Q1E, Evaluation of Stabilit y Data
Mark Shapiro of MCS Pharma Consult- (Geneva, Switzerland, 2003).
tion, USP 37/NF 32, General Chapter USP
ing, Madhavi Mahavadi of BioMarin <1226>, Verification of Compendial Pro-
Pharmaceuticals, and Joshua Ayers of cedures (United States Pharmacopeial
ASQ Solutions. Convention, Rockville, Maryland, 2013,
revised 2018). ABOUT THE CO-AUTHOR
(16) International Council for Harmonisation
References of Technical Requirements for Pharma-
Kim Huynh-Ba
(1) R.G. Hill and H.P. Rang, Eds., Drug Dis- ceuticals for Human Use (ICH) Q2 (R1), is the managing director
covery and Development: Technology in Validation of Analytical Procedures: of Pharmalytik LLC. (www.
Transition (Churchill Livingston, Elsevier, Methodology (Geneva, Switzerland, pharmaly tik.com), which
Edinburgh, Scotland, 2nd Ed., 2012) November 1996, updated 2015). provides consulting ser-
(17) S.W. Baertschi, K.M. Alsante and R.A.
vices in Stability Sciences, Quality
(2) M.W. Dong, Drug Development Process,
Short Course presented at Pittcon, Phila- Reed, Eds., Pharmaceutical Stress Management Systems, and Analytical
delphia, Pennsylvania, March 2019. Testing: Predicting Drug Degradation Development. She is an Adjunct Pro-
(CRC Press, Boca Raton, Florida, 2nd fessor at Temple University’s School
(3) M.W. Dong, LCGC North Am. 33(11),
Ed., 2011). of Pharmacy and Illinois Institute of
764–775 (2015).
(18) H.T. Rasmussen, W. Li, D. Redlich, M.I.
Technology (IIT). Kim is a member of
(4) D. Kou, L. Wigman, P. Yehl, and M.W. the US Pharmacopeia’s Council of
Dong, LCGC North Am. 33(12), 900–909, Jimidar, HPLC Method Development, In
Handbook of HPLC in Pharmaceutical Experts, and the chair of the Chemical
(2015). Medicines Monograph IV Expert Com-
Analysis, S. Ahuja and M. W. Dong, Eds.,
(5) M.W. Dong, HPLC and UHPLC for Practic- (Elsevier, Amsterdam, 2005), Chapter 6. mittee, USP Good Documentation
ing Scientists (John Wiley & Sons, Hobo- Practices Expert Panel, USP Organic
ken, New Jersey, 2nd Ed., 2019), Chap- (19) M.W. Dong and H.T. Rasmussen, HPLC
Method Development Short Course, Impurities of Drug Substance and
ters 9–11. Drug Products Expert Panel. She is on
Eastern Analytical Symposium, Somer-
(6) K. Huynh-Ba, Ed., Handbook of Stabil- set, New Jersey, 2004. the editorial board of AAPS Open, the
ity Testing of Pharmaceutical Products: Journal of GXP Compliance, and the
Regulations, Methodologies and Best (20) R. Maheswaran, Pharm. Technol. 36(5),
1–6 (2012).
Journal of Validation Technology. Kim
Practices (Springer, New York, 2009). has authored ~30 technical publica-
(7) K. Huynh-Ba, Ed., Pharmaceutical Stabil- (21) 21 Code of Federal Regulations (CFR), tions and is the editor of two books on
ity Testing to Support Global Markets Part 211.137, Current Good Manufactur-
stability testing to support registration
(Springer, New York, 2010). ing Practice for Finished Pharmaceuti-
in global markets.
cal Products (US Government Printing
(8) International Council for Harmonisation Office, Washington, DC, 2019).
of Technical Requirements for Pharma-
ceuticals for Human Use (ICH) Q1A (R2), (22) K. Waterman, Understanding and Pre- ABOUT THE AUTHOR
Stability Testing of New Pharmaceutical dicting Pharmaceutical product Shelf life,
Products (Geneva, Switzerland, 2003). In Handbook of Stability Testing of Phar- Michael W. Dong
maceutical Products: Regulations, Meth- is a principal of MWD
(9) International Council for Harmonisation odologies and Best Practices (Springer, Consulting, which provides
of Technical Requirements for Pharma- New York, 2009), Chapter 6. training and consulting
ceuticals for Human Use (ICH) Q8(R2), services in HPLC and UHPLC,
(23) K. Huynh-Ba et al., Meeting Report,
Pharmaceutical Development (Geneva, Analytical Approaches to Ensure Prod- method improvements, pharmaceutical
Switzerland, 2009). uct Quality – AAPS Joint Face-to-Face analysis, and drug quality. He was formerly
(10) International Council for Harmonisation Meeting of the Stability, the Pharmaceu- a Senior Scientist at Genentech, Research
of Technical Requirements for Pharma- tical Impurities, and the CMC Statistics Fellow at Purdue Pharma, and Senior
ceuticals for Human Use (ICH), Q3A(R2), Focus Group, AAPS Open 3 (1), 2017. Staff Scientist at Applied Biosystems/
Impurities in New Drug Substance, https://aapsopen.springeropen.com/ PerkinElmer. He holds a PhD in Analytical
Q3B(R2), Impurities in New Drug Prod- articles/10.1186/s41120-017-0011-z Chemistry from City University of New York.
ucts (Geneva, Switzerland, 2006). (24) H. Williams et al., Pharm. Technol. 41(3), He has more than 100 publications and
(11) International Council for Harmonisa- 52–57 (2017). a best-selling book in chromatography.
tion of Technical Requirements for (25) D. Lavrich, Rapid Development of He is an editorial advisory board
Pharmaceuticals for Human Use (ICH) Robust Stability Models Using Semi- member of LCGC North America and
Q6A, Specifications: Test Procedures Empirical Design Space, AAPS Webi- the Chinese American Chromatography
and Acceptance Criteria for New Drug nar, March 24, 2016. https://aapsopen. Association. Direct correspondence to:
Substances and New Drug Products, s pring eropen.com /ar ticles/10.118 6/ LCGCedit@MMHGroup.com
(Geneva, Switzerland, 1999). s41120-017-0018-5#article-info
338 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
FOCUS
ON
BIOPHAR M ACEU TIC AL
ANALYSIS
T he quality of a biotherapeutic is
defined by its various biological,
immunological, structural, and physi-
of large proteins may generate more
than 100 peptides. Given the random
distribution of peaks, even a system
Based on how the eluate is trans-
ferred from one dimension to another,
MD separations can be classified as
cochemical attributes. Some of these with a peak capacity of 10,000 would on-line and off-line. The MD system is
are post-translational modifications be able to resolve only 98% of the mix- said to be on-line when the eluate of
(PTMs), other product-related het- ture of 100 tryptic peptides (6). Thus, one dimension is directly transferred
erogeneities, process related impu- in the case of biopharmaceuticals, to the next dimension, and separation
rities, and host cell related impuri- which are mostly large molecules, use occurs in real time in both the dimen-
ties. An array of different techniques of one-dimensional ( 1D) chromatogra- sions. On the other hand, in the case
contributes to the analytical toolbox. phy usually does not yield complete, of off-line systems, eluates from the
These are utilized for characterization resolved product coverage (6,7). first dimension are collected as frac-
of the biotherapeutic product, and Multidimensional (MD) separation tions, and then later subjected to sep-
especially the myriad of other species systems, in which two or more sepa- aration in the next dimension (9,12).
that are also present in the final drug ration methods are coupled, can be Both approaches have some advan-
product (1–3). One of the most com- used in such cases to increase the tages and disadvantages (Table I).
monly used analytical techniques for peak capacity of the system. This will Additionally, MD approaches can
routine process evaluations and qual- then allow the complete resolution, it also be classified as heart-cutting or
ity control in the industry is reversed- is hoped, of all constituent analytes comprehensive ( 1 D× 2 D), depending
phase liquid chromatography-mass (8). These systems improve sample on the part of the 1 D elution con-
Icon Image: (Columns) Joe Zugcic / Zugcic Photographers, Inc.
spectrometry (LC–MS/MS). Due to its resolution manifold by first resolving tent that is transferred in the second
robustness, reproducibility, and easy the components in the first dimension. dimension. In the heart-cutting MD
coupling with the MS, LC–MS has The eluate is then transferred and fur- approach, only a part of the 1 D elu-
become the state-of-the-art technol- ther resolved in the second or subse- ate is subjected to separation in the
ogy for routine characterization of quent dimensions (9). Such a separa- next dimension. However, in a com-
biotherapeutics in the industry (3,4). tion was first reported with the use of prehensive approach, the whole elu-
However, LC-based separations suf- two-dimensional (2D) paper chroma- ate of the 1 D step is subjected to a
fer from a significant limitation aris- tography in 1944 (8). Since then, differ- 2D separation (9).
ing from incomplete separation and ent separation techniques have been Whether on-line or off-line, an MD
co-elution of target analytes (5). One coupled in a two-dimensional or mul- separation system is useful if it fulfills
reason for lack of baseline resolution tidimensional manner, and have been two fundamental requirements, as
is that LC can generate peak capaci- proven useful for the analysis of com- suggested by Giddings (8). First, the
ties up to 1000, and proteolytic digest plex biological samples (9–11). separation techniques in the MD set
WWW.CHROMATOGRAPHYONLINE.COM JUNE 2020 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 339
up must be orthogonal. This means TABLE I: Advantages and disadvantages of on-line and off-line coupling
that the resolving principles of the
Parameter On-line Off-line
separation techniques must be based
on different chemical or physical prop- Speed Fast (12) Slow (12)
erties of the analytes. Thus ideally, for Automation Feasible (10) Difficult (10,13)
a given analyte, the peak distribution Mobile phase compositions Limited options (12) More options (12)
in the two dimensions must be uncor- Optimization Difficult (12) Easy (12)
related. The second principle states Sample manipulation
that the resolution of the first dimen- Not allowed (10) Possible (10)
between dimensions
sion must not be lost during separa- Resolution Less (9) More (9)
tion in subsequent dimensions (8,12).
Multidimensional LC–LC Coupling been combined. Such couplings have applied for the characterization of bio-
One of the most successful MD sys- been successfully implemented for the therapeutics, are discussed below. In
tems, widely accepted by both aca- analysis of biopharmaceuticals (11,15). all these methods, reversed-phase LC
demia and industry alike, is that of However, not all LC modes can be has been used for the second-dimen-
2D–LC-based separations (14). These combined with each other, due to sol- sion separation, likely because of the
separations involve separation of com- vent immiscibility and other compat- high-quality reversed-phase LC col-
pounds at two different conditions, ibility issues (16). These 2D-LC systems umns available and their compatibility
using either different modes of LC or have been widely accepted by the bio- with MS (34).
using a single-mode operated at two pharmaceutical industry. Some of the
different conditions. Many LC modes, applications of these techniques in the Reversed-phase LC x
such as ion exchange chromatography industry are summarized in Table II. Reversed-phase LC Coupling
(IEC), size exclusion chromatography Some of the most commonly used, This separation system involves resolv-
(SEC) and reversed-phase LC, have 2D-LC methods, which have been ing the analytes via reversed-phase
ADVANTAGE
See What It Can Do for You and Your Lab
Pure Chromatography
340 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
concentration at 16 ppm. To decrease tic peptides of trastuzumab. Although phase LC in the second dimension,
analysis time and improve peak shape, the system offers high orthogonal- as they are eluted with the change in
the second dimension was operated ity, concerns persist due to the pres- the pH or ionic concentration of the
at a high flow rate and high tempera- ence of a high organic content in the mobile phase. The advantage of using
ture, which then led to a 15% loss in mobile phases (35). this system is its ease of use, as well
peak capacity as compared to that as the minimal sample manipulation
obtained with standard LC conditions SEC x Reversed-Phase LC Coupling requirements. Both cation-exchange
(27). Limited orthogonality is, how- Size-exclusion chromatography (SEC) x and anion-exchange chromatography
ever, a major issue of this approach reversed-phase LC separates analytes have been coupled to reversed-phase
(36). Column stability and possible on the basis of size and hydrophobic- LC to separate protein mixtures and
erosion of fused silica tubes at high ity in the first and second dimensions, tryptic digests (3,37). For example,
pH are some of the other factors that respectively. The system is thus highly rapid analysis of charge variants of
must be considered before setting up orthogonal, and it is useful for peptide mAbs without much sample manipu-
this separation system (35). separations. Synthetic polymers and lation was performed, with a generic
biopolymers are some of the major 2D-LC–MS method. Two independent
HILIC x Reversed-Phase LC Coupling application areas of this system (37). gradients were used on a single chro-
One of the best orthogonalities dem- A 2D-LC–MS based, semi-automated matographic system, consisting of
onstrated is exhibited by the hydro- platform was developed for charac- two independent column compart-
philic interaction chromatography terization of impurities in monoclonal ments, with built-in switching valves.
(HILIC) x reversed-phase LC combina- antibodies, with a week-long data Reversed-phase trap cartridges were
tion (36). In the first dimension, HILIC collection time. Breakdown products used in the second dimension for
separates analytes mainly based on were identified in stress induced mAb desalting first dimension fractions
their separation between the mobile samples, using SEC–reversed-phase before MS analysis. These cartridges
phase and water-enriched layer LC–MS/MS analysis. The low molecu- enabled fast conditioning and re-
adsorbed onto the stationary phase. lar weight breakdown products were equilibration of columns, with minimal
In the second dimension, reversed- enriched by fractionation via SEC in backpressure, during the switching of
phase LC separates them on the basis the first dimension, and the fractions columns. The poorly resolved and low
of their hydrophobicity. The use of this were resolved via C8 chromatography abundance peaks of IEC in the first
setup has been recently reported in a in the second dimension. Both on-line dimension, were preconcentrated on
study in which a commercially avail- and off-line mass spectrometry analy- the reversed-phase traps using mul-
able 2D-LC system was used for the ses were performed, using the C8 elu- tiple injections. Using an in-process
bottom-up analysis of trastuzumab. ate. A part of the C8 eluate was thus sample of a mAb, fractions of interest
A novel dual loop interface was used directed to the MS while a major part were collected from IEC, and then sub-
for transferring the samples between was collected as fractions. While on- sequently analyzed via reversed-phase
the two dimensions. The use of high line analysis was used for intact mass LC, combined with an on-line elec-
elution strength and a high volume of analysis, the off-line set up included trospray ionization (ESI)-time-of-flight
injection solvent, however, led to com- analysis under both reducing, as well (TOF-MS) based detection. Chain spe-
plete or partial elution of polar pep- as non-reducing conditions, so as to cific information about mAbs was also
tides in the void volume. Several pre- elucidate impurity structure. Various deduced by a disulfide reduction step.
cautions, such as high flow rate and low anticipated products, as well as unpre- This was incorporated as an optional
temperature of the second dimension, dicted modifications, such as frag- step into the workflow in an on-line
were taken to minimize peptide break- ment antigen-binding (Fab) fragments manner. Contamination, sample loss,
through in reversed-phase LC, but it and hydrolyzed, free light chain, were and degradation were minimized with
could not be completely eliminated. identified. The approach was also pro- the use of an on-line setup. The whole
Unfortunately, the HILIC x reversed- posed to be useful for charge variant system was fully automated to allow
phase LC combination was challeng- and aggregation analysis of mAbs. for high throughput analysis, minimum
ing, and offered limited coverage of sample handling, and rapid analysis.
the separation space as well. The study IEC x Reversed-Phase LC Coupling The system was shown to be useful
compared HILIC x reversed-phase LC In this mode of duality HPLC, ion for charge, sequence, and chemically
with strong cation exchange (SCX) exchange chromatography (IEC) is unstable mAb variant analysis. Minor
x reversed-phase LC and reversed- used in the first dimension, which glycoforms were also identified at the
phase LC x reversed-phase LC, and it separates molecules based on their intact protein level, due to the sepa-
found that the other two techniques charge. Analytes are transferred from ration of charged isoforms in the first
are superior for the separation of tryp- IEC in the first dimension to reversed- dimension by IEC (20).
342 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
400
300
Multidimensional LC-CE Coupling
Due to the limited orthogonality and
200
other disadvantages associated with
100
LC modes in the second dimension,
0 alternative separation techniques,
20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
Time (min) having different separation principles,
(b) x10 7 have been attempted to improve the
1
3
2.8 separation further. Capillary electro-
2.6
2.4
phoresis (CE) is an attractive alterna-
Counts (x107)
2.2
2
1.8 tive in this regard, because it separates
1.6
1.4 analytes based on their charge to size
1.2
1
0.8
ratios, complementary to all modes
0.6
0.4 of LC (10,39). Electrophoresis with
0.2
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
chromatography has been described
Acquisition Time (min) as a powerful 2D-separation method
by Giddings (9). The use of CE as the
second dimension in 2D systems can
FIGURE 1: Off-line LC-CE-MS of a protein digest. A separation of a protein digest was performed thus provide high efficiency and rela-
using reversed-phase LC in the first dimension and capillary electrophoresis (CE) in the second tively fast separations as compared to
dimension. (a) Separation profile of the digest with reversed-phase LC in the first dimension. Sol- LC-based methods. However, special
vent A and solvent B were 0.1% formic acid in water and 0.1% formic acid in acetonitrile, respec-
tively. Overlapping peaks at 38–39 min are circled in green. (b) CE–MS profile of the fraction,
coupling techniques are required for
containing the peaks circled in green in panel a, in the second dimension. CE separation was CE based setups, due to the use of
performed in a 56 cm long polyvinyl alcohol coated capillary at 30 kV. Background electrolyte was small volumes and high electric fields
2% acetic acid and detection was performed at 214 nm. (13). The history of coupling chroma-
tography and electrophoresis dates
The coupling of SCX in the first or less correlated selectivity, and a back to 1948, and since then the cou-
dimension and reversed-phase LC in lower than maximum peak capacity pling has been shown to improve the
the second is the most widely used is usually achieved (10,36). Another resolution of various biological and
2D-LC system, especially in bottom- major limitation of the 2D–LC system environmental samples (9).
up proteomics-based studies. Ana- is that the separation processes have Various methods based on LC–
lytes are separated on the basis of longer analysis times. Various meth- CE, CE–LC, and SEC–CE have been
their charge in the first dimension, ods to speed up the LC-based separa- reported in the literature, primarily for
and hydrophobicity in the second tions via altering column length, tem- proteome and peptide level analysis.
dimension (35). This method, however, perature, and particle size have been An in-depth review of the applica-
suffers from a significant limitation reported. However, each of these tions of different modes of CE in MD
in terms of the peak capacity in the methods has its own limitations (37). In separations has been recently pub-
first dimension. The SCX chromato- 2D-LC, for example, to achieve faster lished (13). The ability of CE to resolve
gram is sparsely populated with posi- separation, the second dimension LC LC peaks has been discussed in the
tively charged tryptic peptides, most column is run at high flow rates, which literature (40), and is demonstrated
of which (+2 and +3 charged pep- compromises resolution. The second in Figure 1. In this work, the tryptic
tides) tend to be eluted close to each dimension can also be made faster digest of a protein was resolved using
other. Another major limitation of this by operating the column at high tem- reversed-phase LC in the first dimen-
method is the limited orthogonality of perature, which reduces the viscosity sion, then fractionated, with each
the system (about 50%). Furthermore, of the mobile phase, and allows the fraction subjected to CE–MS analy-
hydrophobic peptides may be eluted user to increase the eluent flow rate, sis in the second dimension. Several
with a poor peak shape, or may be but this may affect the stability of overlapping peaks were observed in
incompletely recovered with SCX (36). the stationary phase and the analyte LC that were efficiently resolved by
The major limitation with these sys- (38). Thus, the use of high tempera- CE in the second dimension (Figure
tems, however, is that they show more ture and other operating parameters, 1). For instance, the peaks from 38
WWW.CHROMATOGRAPHYONLINE.COM JUNE 2020 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 343
to 39 min (circled in green in Figure mixture, immunoconjugates of a mAb, was also useful for obtaining a deeper
1a), remained unresolved in reversed- and disulfide isomer heterogeneity of understanding about the conjugation
phase LC, as observed in the corre- IgG2-mAbs (9,10,40–43). For example, process. Although this method was
sponding UV chromatogram. CE–MS a size exclusion chromatography- well suited for analyzing reduced and
of the corresponding fraction showed imaged capillary isoelectric focusing denatured mAbs, and offered several
baseline resolution of the same peaks (SEC-iCIEF) method was proposed for advantages over 2D PAGE, it required
(Figure 1b). the quantitative charge heterogeneity large amounts of protein sample
Although these reports clearly analysis of heavy and light chains of a for analysis (28).
demonstrate the utility of LC–CE in mAb. Over 10 different mAbs, includ-
biotherapeutics, LC–CE suffers from ing both IgG1 and IgG4 subtypes, Four-Dimensional (4D) HPLC-MS
significant limitations in terms of injec- were characterized using this platform. MD setups are not limited to just two-
tion and elution volume disparities, The platform was found to be less dimensional analysis. Use of more
between the LC and CE dimensions. time-consuming, less labor-intensive, than two dimensions is also emerg-
There is also the need to isolate CE more reproducible, and allowed fully ing for the analysis of biotherapeu-
from the other parts of the system, due quantitative analysis of charge variants tics. For instance, recently, an on-line
to its high electric field. Due to these compared to the traditional 2D poly- 4D HPLC-MS approach was reported
reasons, LC–CE has still not gained as acrylamide gel electrophoresis (PAGE) for characterization of mAb variants
much attention as 2D-LC (9). A com- analysis. Results were obtained in a sin- (44). The mAb sample was resolved
parative analysis of the advantages gle day, but only with automated SEC and fractionated using IEC in the first
and disadvantages of the commonly and iCIEF platforms. The method was dimension. These fractions were then
used MD methods is given in Table III. also used for characterization of mAbs transferred to the 2D reversed-phase
Nevertheless, the coupling of LC with under accelerated stability conditions, column, where they were reduced
CE has been described for analysis and it was proposed to be useful for and washed. The reduced mAb chains
of natural compounds, peptide frag- generating simplified electrophero- were then eluted onto the third col-
ments, tryptic peptides of a protein grams for conjugated antibodies. It umn, where they were digested.
344 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
The digested peptides were then detected using LC–MS in development, and data analysis strategies in the future are
the fourth dimension. The whole system was fully automated expected to make it a routine characterization technique for
from fractionation to peptide mapping. Prefractionation by quality control processes in the industry. Various modes are
IEC allowed reduction and digestion of separated charge available for 2D-LC systems, but these have limited orthogo-
variants, resulting in higher signal intensity and sensitivity. nality and peak capacity. Recent years have witnessed the
The coupling of MS/MS allowed for localizing the modifica- growth of CE as a method of choice for MD separation sys-
tions at the amino acid level. The system was highly sensi- tems. Although these are not yet as widely used as 2D LC-
tive, and it was able to detect deamidation present, even at based methods, interest in their deployment is growing rap-
0.5% of the main peak. Five out of six oxidation sites were idly. They represent one of the major developments in the
identified in the study. The system, however, showed lim- industry which should become more obvious in the future.
ited retention of small and hydrophilic peptides. Although
the authors presented IEC in the first dimension, the 4D References
set-up was designed so that SEC could be coupled in place (1) A.J. Reason, A. Weiskopf, and A.S. Rathore, BioPharm Intl. 27(7),
of the IEC for the characterization of size variants. 1–8 (2014).
(2) A.S. Rathore, I.S. Krull, and S. Joshi, LCGC North Am. 36(6),
Conclusions 376–384 (2018).
Biopharmaceuticals are complex molecules, and they (3) A.S. Rathore, I.S. Krull, and S. Joshi, LCGC North Am. 36(11),
814–822 (2018).
demand complex, analytical procedures with maximum
(4) M. Pioch, S.C. Bunz, and C. Neusüß, Electrophoresis 33(11),
resolving power. To meet this need, one-dimensional analyti- 1517–1530 (2012).
cal methods are not sufficient for achieving complete analy- (5) S.K. Grebe and R.J. Singh, Clin. Biochem. Rev. 32, 5–31 (2011).
sis. Multidimensional systems offer an alternative, promis- (6) G. Vanhoenacker, I. Vandenheede, F. David, P. Sandra, and K.
ing approach in this regard. 2D-LC is commonly used as a Sandra, Anal. Bioanal. Chem. 407(1), 355–366 (2015).
multidimensional separation system for biopharmaceutical (7) J.R. Lill, Analytical Characterization of Biotherapeutics (John
analysis, and it has now been accepted by many biophar- Wiley & Sons, Hoboken, New Jersey, 1st Ed., 2017) pp. 1-14.
maceutical companies. Simplified instrumentation, method (8) W. Grochocki, M.J. Markuszewski, and J.P. Quirino, Electropho-
resis 36(1), 135–143 (2015).
(9) L. Ranjbar, J.P. Foley, and M.C. Breadmore, Anal. Chim. Acta
950, 7–31 (2016).
(10) P. Cesla, P. Dinisová, and J. Fischer, Sens. Electroanal. 6, 357–
368 (2011).
(11) B.W.J. Pirok, D.R. Stoll, and P.J. Schoenmakers, Anal. Chem.
91(1), 240–263 (2019).
(12) C.R. Evans and J.W. Jorgenson, Anal. Bioanal. Chem. 378, 1952–
1961 (2004).
(13) F.J. Kohl, L. Sánchez-Hernández, and C. Neusüß, Electrophore-
Over 25 Years of sis 36(1), 144–158 (2015).
(14) X. Wang, S. Buckenmaier, and D. Stoll, J. Appl. Bioanal. 3(5),
Quality HPLC and 120–126 (2017).
(15) H. Wang and S. Hanash, J. Chromatogr. B 787, 11–18 (2003).
SFC Columns (16) P.W. Carr and D.R. Stoll, Two-dimensional Liquid Chromatogra-
phy (Agilent Technologies, 2015).
(17) R.E. Birdsall, H. Shion, F.W. Kotch, A. Xu, T.J. Porter, and W.
Chen, MAbs 7(6), 1036–1044 (2015).
(18) S. Schneider, Online 2D-LC Characterization of Monoclonal
Antibodies with Size Exclusion and Weak Cation Exchange
Chromatography (Agilent Technologies, 2016).
(19) Y. Li, D. Hewitt, Y. K. Lentz, J. A. Ji, T. Y. Zhang, and K. Zhang,
Anal. Chem. 86(10), 5150–5157 (2014).
(20) M. Alvarez, G. Tremintin, J. Wang, M. Eng, Y. Kao, J. Jeong, V.T.
Offering a wide range Ling, and O. V Borisov, Anal. Biochem. 419(1), 17–25 (2011).
of phases and dimensions for (21) J.J. Gilroy and C.M. Eakin, J. Chromatogr. B 1060, 182–189 (2017).
both analytical and prep (22) M.T. Mazur, R.S. Seipert, D. Mahon, Q. Zhou, and T. Liu, AAPS J.
14(3), 530–541 (2012).
(23) Y. Li, C. Gu, J. Gruenhagen, K. Zhang, P. Yehl, N.P. Chetwyn, and
C.D. Medley, J. Chromatogr. A 1393, 81–88 (2015).
www.pci-hplc.com | 609.860.1803
(24) D.R. Stoll, D.C. Harmes, J. Danforth, D. Guillarme, S. Fekete, and
A. Beck, Anal. Chem. 87(16), 8307–8315 (2015).
WWW.CHROMATOGRAPHYONLINE.COM JUNE 2020 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 345
(25) K. Sandra, A. Ortiz, and P. Sandra, LCGC (35) F. Xie, R.D. Smith, and Y. Shen, J. Chro- ABOUT THE CO-AUTHOR
Europe 28(10s), 45–48 (2015). matogr. A 1261, 78–90 (2012).
(26) J.H. Thompson, W.K. Chung, M. Zhu, L. (36) M. Gilar, P. Olivova, A.E. Daly, and J.C.
Ramesh Kumar is a grad-
Tie, Y. Lu, N. Aboulaich, R. Strouse, and Gebler, Anal. Chem. 77(19), 6426–6434 uate student at the Depart-
W.D. Mo, Rapid Commun. Mass Spec- (2005). ment of Chemical Engineer-
trom. 28(8), 855–860 (2014). ing at the Indian Institute of
(37) D.R. Stoll, X. Li, X. Wang, P.W. Carr,
Technology in Delhi, India.
(27) C.E. Doneanu, A. Xenopoulos, K. Fad- S.E.G. Porter, and S.C. Rutanb, J. Chro-
gen, J. Murphy, S.J. Skilton, H. Prentice, matogr. A 1168(1–2), 3–43 (2007).
M. Stapels, and W. Chen, MAbs 4(1), (38) M. Kivilompolo, J. Pó, and T. Hyö-
24–44 (2012). tyläinen, LCGC Europe 24(5), 232–243 ABOUT THE COLUMN EDITOR
(28) R.P. Vanam, M.A. Schneider, and M.S. (2011).
Marlow, MAbs 7(6), 1118–1127 (2015). (39) D. Chen, X. Shen, and L. Sun, Analyst Ira S. Krull is a Professor
(29) R.K. Rober t s, H.J. Holovic s, M.R. 142(12), 2118–2127 (2017). Emeritus with the Depart-
Thompson, and C.W. Demarest, Elec- (40) R. Aebersold and H. D. Morrison, J. ment of Chemistry and
trophoresis 31(3), 448–458 (2010). Chromatogr. 516, 79–88 (1990). Chemical Biology at North-
(30) J. Liu, H. Zhao, K.J. Volk, S.E. Klohr, E.H. eastern University in Bos-
(41) S.E. Rudnick, V.J. Hilser, and G.D.
Kerns, and M.S. Lee, J. Chromatogr. A Worosila, J. Chromatogr. A 672(1–2),
ton, Massachusetts, and a member of
735, 357–366 (1996). 219–229 (1994). LCGC’s editorial advisory board.
(31) R. Kumar, R.L. Shah, and A.S. Rathore, (42) H.J. Issaq, K.C. Chan, C. Liu, and Q. Li,
J. Chromatogr. A (2020). doi:10.1016/j. Electrophoresis 22(6), 1133–1135 (2001).
chroma.2020.460954
(43) E. Tamizi and A. Jouyban, Electrophore-
(32) J. Hühner, K. Jooß, and C. Neusüß, Elec- ABOUT THE COLUMN EDITOR
sis 36, 831–858 (2015).
trophoresis 38(6), 914–921 (2017).
(44) C. Gstöttner, D. Klemm, M. Haberger, A. Anurag S. Rathore is a
(33) K. Jooß, J. Hühner, S. Kiessig, B. Moritz, Bathke, H. Wegele, C. Bell, and R. Kopf, professor in the Depart-
and C. Neusüß, Anal. Bioanal. Chem. Anal. Chem. 90(3), 2119–2125 (2018). ment of Chemical Engineer-
409(26), 6057–6067 (2017). (45) C. Song, M. Ye, G. Han, X. Jiang, F. ing at the Indian Institute of
(34) K. Sandra and P. Sandra, Bioanalysis Wang, Z. Yu, R. Chen, and H. Zou, Anal. Technology in Delhi, India.
7(22), 2843–2847 (2015). Chem. 82(1), 53–6 (2010).
Benson polymeric TM
bensonpolymeric.com
775.356.5755
40419080306_BENSON
magenta
cyan
yellow
black POLYMERIC_pgxx_HPH_1-1.pgs 07.23.2019 20:56 ADVANSTAR_PDF/X-1a
346 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
Gone is the burdensome FDA require- even if it is to get “proper” integration. Sum- not to describe the principles of peak integra-
ment for a manager to preapprove any marizing these citations, we have the follow- tion, as this can be found in the excellent book
manual integration (21), but this is replaced ing areas of non-compliance: by Normal Dyson (23), or in tutorials and man-
by control via a plan or procedure, together • Integrating into compliance uals from CDS suppliers. A suggested peak
with a listing of chromatograms requiring • Failure to retain integration parameters integration workflow is shown in Figure 1 (2,
manual integration (moving baselines) and (for example, a lack of complete data) 24), and we will discuss various aspects of this
the before and after chromatograms. For a • Lack of a procedure for manual integration workflow throughout this article.
study involving the analysis of >1000 samples, • Inappropriate use of integrate inhibit
this could be quite a large burden especially • Unscientific integration Why Manually Integrate Peaks?
if the elimination phase, is relatively long and • No quality oversight We need to consider why we need to inte-
the drug is slowly metabolized or excreted. In my view, the FDA are wrong in their grate peaks manually. The reason is that
These requirements emphasize the need to requirement for a procedure on manual inte- there are situations where a CDS cannot inte-
have robust and reliable analytical proce- gration. To avoid these issues, we must have grate peaks correctly, and some examples of
dures, as discussed in an earlier “Data Integ- a procedure that is structured and compliant, this are:
rity Focus” article (22). and a scientific approach to peak integra- • Split peaks
tion as a whole. What is required is a holistic • Shoulder peaks
A Parcel of Rogues approach with an SOP for all chromatographic • Tailing peaks
A small selection of FDA 483 observations integration, of which manual integration is an • Baseline noise
and warning letter citations of chromato- important subset. Further information about • Negative peaks
graphic peak integration failures are pre- an integration SOP, including banned prac- • Co-eluting peaks
sented in Table I, with the key problems tices such as peak skimming or enhancing, • Rising or falling baselines
highlighted in bold. One piece of free advice can be found in the articles by McDowall, • Slowly eluting peaks (where the CDS
I would offer is not to say to an inspector that Newton and McDowall and Longden and has difficulty identifying the peak start
chromatographers “play with parameters,” McDowall (2–4). The purpose of the SOP is and end).
TABLE I: A Selection of FDA 483 observations and warning letter citations on peak integration
You failed to establish adequate test procedures. For example, your analyst manually integrated a HPLC
Hubei Danjiangkou Danao
test for <redacted> API despite the fact that the chromatogram lacked peak resolution.....You lacked an
Pharmaceutical Co Ltd (30)
approved protocol for manual integration or quality oversight of the practice.
Your firm failed to properly integrate co-eluting peaks during impurity testing of phentermine HCL
capsules, which resulted in your analysis failing to detect out-of-specification (OOS) results for at least
KVK-Tech, Inc (31)
one lot of drug product.....Multiple examples of your firm’s failure to properly integrate closely
co-eluting peaks were observed during our inspection.
The reasons for the inability of the integra- tions to this statement. None, repeat none. published in 2018, formally defined manual
tion method may be due to: If the peak integration for the sequence integration as:
• Poor method development and valida- is acceptable, then the reportable results
tion where the analytical procedure or can be calculated and reviewed; high fives “(The) (p)rocess used by a person to modify
the integration method is not optimized all around! However, if the peak integra- the integration of a peak area by modifying
or robust. tion is not acceptable we come to the the baseline, splitting peaks or dropping a
• Complex sample matrices resulting in first decision point: Is manual integration baseline as assigned by the chromatogra-
interfering peaks that may still be present allowed for this analysis? If not, we trigger phy software to overrule the pre-established
after sample preparation, such as biologi- a laboratory investigation. integration parameters within the chromato-
cal samples, contrast media, and fermen- The big problem with the citations about graphic software (19).”
tation media. the lack of a manual integration procedure
• Complex mixtures of similar analytes in Table I is that there is no definition of the Is this definition acceptable?
These situations may result in a heavy term “manual integration.” • It is wordy, repetitious, and could be
manual integration workload, as the better phrased.
CDS method is not able to integrate all Defining Manual Integration • The use of the word “overrule” is conten-
peaks automatically. In a “Questions of Quality” column on tious. Chromatography is a dynamic pro-
integration, it was noted that there was cess, and a CDS application can struggle
A Suggested Peak no definition of manual integration, and it to separate overlapping peaks which are
Integration Workflow: 1 was implicitly defined as manual placement obvious to a trained eye.
The first part of the integration workflow in of the baseline by a chromatographer (2). • The biggest issue with this definition
Figure 1 is that all peak integration in the first Mark Newton and I reiterated this defini- is that there is no mention of scien-
instance must always be conducted using tion in our “Data Integrity Focus” article in tific soundless, as defined in the FDA
automatic integration. There are no excep- 2018 (3). The PDA’s Technical Report 80, also GMP regulations (5).
WWW.CHROMATOGRAPHYONLINE.COM JUNE 2020 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 349
Chromatographic Run
Automatic
integration
Manual Complete
Laboratory
No Integration No Acceptable? Yes Analysis
Investigation
OK?
Yes
Manual
Integration
Options
1. Rename
peaks or adjust
windows 2. Adjustment of
integration
parameters 3. Manual
reposition of
baselines
Manual Manual
Intervention Reintegration
Therefore, a simpler, more concise, and side auditors recognize the legitimate versus
better definition of manual integration egregious use of integration?
should be “manual repositioning of peak Banning the use of manual integration is a
baselines with scientific justification for common response to avoid questions about
their positioning.” data integrity. However, there are three out-
Implicit within this definition is the use of comes to this crude (and stupid) action:
CDS software; otherwise, you’d be draw- • Laboratories will have to accept poor and
ing baselines on paper. However, this also inconsistent integration
requires that the chromatographer is trained, • Analysts will find a workaround that per-
and, ideally, software technical controls mits them to integrate each run with a
should prevent manual repositioning of different set of integration parameters
baselines where this is not justified by the (typically involves performing quanti-
type of the analysis. fication in a LIMS, or worse, a spread-
sheet, without traceability back to the
Should Manual integration methods)
Integration Be Banned? • Analysts will be forced to spend hours
From the citations in Table I, would it be rea- of their day developing complex and
sonable to ban manual integration in regu- manipulative methods to address varia-
lated laboratories? Let’s think this through. tions between chromatograms with a
Experienced analysts know that chromato- single processing method. Typically, this
graphic analysis can be affected by tempera- will require many “integration events”
ture, humidity, and column history, as well that could even include placing peak
as mobile phase preparation, such that one starts and ends at specific time points;
day’s analysis often varies slightly from the in effect, performing manual integration
previous day’s run. To achieve consistent out- under the guise of an automated method
put and measurement, it is critical to adapt to deceive a reviewer (4).
or optimize factors such as peak detec- Consider, also, the nature of an analysis.
tion threshold or retention time windows My experience is based on analysis of small
to ensure consistent, correct, and accurate heterocyclic molecules; however, I acknowl-
peak identification and measurement (4). But edge that there are biologicals, macro-
how can reviewers, approvers, quality, or out- molecules, and chiral separations that will
350 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
tific judgment that needs to be backed up FIGURE 2: Scope and order of chromatographic integration (3).
by the procedures within the integration
SOP and the changes retained in the audit This places responsibility on any labora- • Standards injections
trail of the CDS. Table II shows suggestions tory to develop robust analytical chromato- • Other checks, such as quality control sam-
for when manual integration is and is not graphic procedures with reliable separations ples and blanks
allowed, depending on the nature of the that are fit for use as discussed in a previous • Samples.
analysis undertaken by a laboratory. “Data Integrity Focus” article (22). A key com- If any of these items fail acceptance crite-
ponent of this approach is that the resultant ria, STOP! Do not review anything involving
Why Use Automatic Integration? peak integration must be consistent. This is a samples until assay criteria are met. Looking
This is a very simple question to answer from subtle, but vital, difference that is not always at sample results prior to accepting the run
a business perspective: sequences where appreciated. could be considered “integrating into com-
integration is automatic are very much pliance,” as you wanted to invalidate testing
quicker to review. In contrast, manual inte- Sequence of Data File Integration you didn’t like (or accept one you did like).
gration is slow, especially for sequences with In 2018, Mark Newton and I discussed the Fresenius Kabi Oncology received a warning
a large number of injections with complex order of integration and processing of any letter from the FDA for such behavior (32).
separations where each injection must be injection sequence (3), and, due to the
integrated manually. Often the review can importance of correct integration, it is worth Is the FDA Inhibiting Integration?
take longer than the actual analysis (3). Con- repeating here. Figure 2 shows the order of The use of inhibit integration function is cur-
sequently, a faster review process means that integration evaluation: rently a hot regulatory topic, as can be seen
analysis is completed quicker, a batch can • Automatic integration and manual from in the FDA regulatory citation of Divi’s
be released sooner, and the organization intervention Laboratories (shown in Table I), where this
cash flow benefits. There is also the bonus of • Manual integration (if permitted) function was used four times in the middle of
lower regulatory scrutiny. • System suitability test injections a chromatogram.
This citation is based on 21 CFR 211.160(b) (5) that was discussed cions that the excluded peaks might be real impurities, exclud-
earlier in this article, and the key questions to ask are if, where ing these system peaks needs to be carefully documented and
and when can integrate inhibit be used. It has been said in some justified in the method development and validation reports;
audits and inspections that this function cannot be used. This is otherwise, they should be integrated and marked clearly as
an untenable situation, as there is no explicit or implicit statement system peaks.
in the GxP regulations for this attitude. However, it comes back to
scientific soundness, and a laboratory must be able to justify the System Evaluation Injections
use of the function. Trial injections using actual samples feature in many warning let-
Let us consider the following scenarios: ters (33) and Question 13 of the FDA Data Integrity Guidance (16),
• There is baseline perturbation with a large negative peak after which states:
an injection.
• A peak of interest elutes shortly after the negative peak. The use of “FDA prohibits sampling and testing with the goal of achieving a spe-
integrate inhibit is fully justified from the start of the injection, void cific result or to overcome an unacceptable result (e.g., testing different
volume, solvent front, and until the baseline has returned to nor- samples until the desired passing result is obtained).”
mal and before elution of the peak of interest. Otherwise, there is
a large probability that baseline placement of the analyte could be This is correct and should never be acceptable in a GxP
adversely influenced by the negative peak. This would also result laboratory SOP.
in more manual integration. However, consider the following situation: You are analyzing low
• Similar scenarios occur when extraneous peaks are washed from volume samples from a non-clinical study. There is a total volume of
the column, or baseline perturbations from mobile phase change 20 µL plasma sample that is extracted, and there is only enough for
during a gradient elution, or wash at the end of a chromatogram. a single injection from each sample. Ask yourself the question: Are
What is more problematic is the use of integrate inhibit in you going to commit an analytical run of samples without checking
the middle of a run, as cited above. If system, blank, or other that the system is ready? The cost of repeating the study is a high
non-sample peaks occur in the middle of a chromatogram, tra- six figure sum if a run does not work. Therefore, from a practical
ditionally those peaks were not integrated. Because of suspi- perspective, we need a way of checking that a system is ready for
analysis, but does not involve testing into compliance with samples.
Ah, somebody says to use SST injections. The problem is that you
may need several replicates to determine if the system is ready, and
SST should never be started until you are confident that the system
is equilibrated.
Consider the following approach for system evaluation or readi-
SERVING ROYALTY. EXCEEDING EXPECTATIONS. EVERY MOMENT. ness injections or equilibration checks (4):
• The ability to use system evaluation injections must be docu-
mented in an applicable SOP or analytical procedure.
• The minimum column equilibration time needs to be documented in
the method to avoid excessive system readiness injections.
• Only system evaluation injections prepared from a suitable refer-
ence standard can be used to evaluate if the chromatographic sys-
tem ready. Records of the solution preparation must be available.
Ideally, a test mixture which mimics the separation characteristics,
but is easily distinguishable from real samples could be used.
• Should the maximum number of system evaluation injections
that can be made be documented in the procedure before a
problem with the chromatographic system needs to be investi-
• Provider of top brand HPLC instrumentation products
gated? If the cause is thought to be an equilibration issue, waiting
• Equivalent to corresponding OEM products
and injecting again should be sufficient. If the system continues
• Serving customers for over 35 years
• Reduce product repair expenses by up to 30%
to not behave, then an investigation is needed, the cause should
• Lifetime Warranty on manufacturing defects be found, remediated, and documented in the instrument log
book before checking the system evaluation again. If the prob-
www.sciencix.com 800.682.6480 sales@sciencix.com
lem requires maintenance to resolve it (for example, a pump
seal replacement), then requalification of the pump should be
conducted and documented before beginning the analysis.
• Using sample preparations must be prohibited in this procedure,
accompanied with staff training.
WWW.CHROMATOGRAPHYONLINE.COM JUNE 2020 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 353
• System evaluation injections are part of that the factors involved in the separation are of manual integration is a regulatory concern,
complete data for the analytical run, and known and controlled adequately (22). Chro- and use needs to be scientifically sound. But
must be included in the instrument log matographic methods should be developed also be aware that, as discussed earlier, man-
book entries, along with any investiga- with automatic integration and the norm ual integration slows down a process, so see
tion and remediation work on the instru- and not the exception. Management need Rule 1 to get the right method depending
ment. A common practice is to store to understand that adequate time must be on the sample matrix and peaks of interest.
the data from these tests in a separate given to method development and validation.
folder or location to the real analyses. This is especially true for pharmacopoeial Rule 4: Understand how the CDS works
This practice needs careful management methods that never work as written. and how the numbers are generated. This
and documentation, as it becomes dif- requires basic training in the principles of peak
ficult to connect those injections to the Rule 2: Never use default integration integration and how a CDS works. The prob-
official laboratory work. Ideally, all work, parameters, always configure specific lem is that with mergers, acquisitions, and
including system evaluation injections, integration for each method. Without encouraging experiences analysts to retire
should be stored in the same location. exception, peak integration and result pro- and employ younger workers, skills are being
cessing must be defined and validated for eroded, and a CDS can be looked at as a
Five Rules of Integration each method so that all peak windows and black box that always gives the right answers.
An integration SOP was discussed earlier. To names are defined, and, if necessary, any
help understand what should be in it and system peaks are identified. Using a default Rule 5: Use your brain – Think! This rule
the associated training, there are five rules of or generic method results in excessive need is sometimes difficult to follow, but it follows
integration to consider (4): for manual integration to name and calculate on from Rule 4. You can have what appears
peaks, and so forth. to be a perfect separation and peak integra-
Rule 1: The main function of a CDS is not tion, but look at peak start and end place-
to correct your poor chromatography. This Rule 3: Always use automatic integration ment—do they look right? Use the zoom
places greater emphasis on the development as a first option and control manual inte- and overlay functions of the CDS to see of
of robust chromatographic procedures so gration practices. Remember that the use standards and samples have the right peak
354 LCGC NORTH AMERICA VOLUME 38 NUMBER 6 JUNE 2020WWW.CHROMATOGRAPHYONLINE.COM
shape, etc. We discussed in the article on Summary agement and Integrity in Regulated GMP / GDP
second person review the need to have Control of peak integration is imperative in a Environments Draft. (Pharmaceutical Inspec-
tion Convention/Pharmaceutical Inspection
an adequate sized monitor; this is a case regulated laboratory. Good peak integration
Cooperation Scheme, Geneva, 2018).
where one is essential (34). The analyst has requires good chromatography. Good chro-
(19) Technical Report 80: Data Integrity Manage-
the responsibility to execute applicable matography requires a well developed and ment System for Pharmaceutical Laboratories
procedures correctly, which includes cor- robust analytical procedure. The bottom line (Parenteral Drug Association [PDA], Bethesda,
rect peak integration. The reviewer, how- is: Are you in control of the analytical proce- Maryland, 2018).
ever, also has a role to ensure that all inte- dure and peak integration? (20) ICH M10 Bioanalytical Method Validation
Stage 2 (International Council on Harmaoni-
gration (whether automated, optimized, sation: Geneva Switzerland, 2019 [draft]).
or manually placed) follows the method References (21) FDA Guidance for Industry: Bioanalytical
guidance for placing baselines as the SOP (1) Able Laboratories Form 483 Observations. Methods Validation (Food and Drug Admin-
describes, especially when the representa- 2005 23 Dec 2019]; Available from: https:// istration: Silver Spring, Maryland, 2018).
www.fda.gov/media/70711/download. (22) R.D. McDowall, LCGC North. Am. 38(4), 233–
tive area for unresolved peaks are being
(2) R.D. McDowall, LCGC Europe 28(6), 336–342 240 (2020).
estimated. Significant peak area manipula- (2015). (23) N. Dyson, Chromatographic Integration
tion should be easily noticed by an experi- (3) M.E. Newton and R.D. McDowall, LCGC Methods (Cambridge: Royal Society of
enced reviewer. North Am. 36(5), 330–335 (2018). Chemistry, Cambridge, United Kindom, 2nd
(4) H. Longden and R.D. McDowall, LCGC Ed., 1998).
Quo Vadis Peak Integration? Europe 21(12), 641–651 (2019). (24) R.D McDowall, Validation of Chromatography
(5) 21 CFR 211 Current Good Manufacturing Data Systems: Ensuring Data Integrity, Meet-
If you think that peak integration is a regula- ing Business and Regulatory Requirements
Practice for Finished Pharmaceutical Prod-
tory issue now, what will it be like in the future? ucts. (Food and Drug Administration: Sliver (Cambridge: Royal Society of Chemistry,
The May 2019 supplement to LCGC Europe Spring, MD, 2008). Cambridge, United Kindom, 2nd Ed., 2017).
gives an interesting glimpse of the future (6) ICH Q7(R1) - Good Manufacturing Prac- (25) FDA Warning Letter Unimark Remedies,
tice for Active Pharmaceutical Ingredients. India, (Food and Drug Administration: Silver
in which Wahab and associates (35) discuss Spring, Maryland, 2014).
(International Conference on Harmonisation:
advanced signal processing techniques that Geneva, 2016). (26) FDA Warning Letter: Divi’s Laboratories Ltd.
could be used in chromatographic integra- (7) EudraLex - Volume 4 Good Manufactur- (Unit II) ( Warning Letter 320-17-34). (Food
tion. The techniques listed are: ing Practice (GMP) guidelines, Part 2 - Basic and Drug Administration: Silver Spring,
Requirements for Active Substances used as Maryland, 2017).
• Deconvolution of extra column effects
Starting Materials (European Commission, (27) FDA Warning Letter Leiner Health laborato-
by Fourier transformation for removing Brussels, Belgium, 2014). ries (Food and Drug Administration: Rockville,
band broadening. (8) R.D. McDowall, LCGC North. Am. 37(4), 265– Maryland, 2006).
• Peak area extraction by iterative curve 268 (2019). (28) FDA 483 Observations: Kashiv BioSciences.
fitting for partial overlapping peaks in (9) R.D. McDowall, Spectroscopy 28(4), 18–25 (Food and Drug Administration: Silver Spring,
(2013). Maryland, 2018).
a chromatogram.
(10) R.D. McDowall, Spectroscopy 31(11), 18-21 (29) FDA Warning Letter: Magafine Pharma Ltd.,
• Model free approaches for peak infor- US FDA:, Silver Spring, Maryland, 2017)
(2016).
mation extraction, another approach for (30) FDA Warning Letter: Hubei Danjiangkou
(11) R.D. McDowall, Spectroscopy 33(12), 8–11
extracting peak areas from overlapping (2018). Danao Pharmaceutical Co., Ltd. (Food and
peaks in complex matrices. Drug Administration, Silver Spring, Mary-
(12) EudraLex - Volume 4 Good Manufacturing land, 2017).
• Direct resolution by Power Law increases Practice (GMP) Guidelines, Chapter 4 Docu-
mentation (European Commission, Brussels, (31) FDA Warning Letter: KVK-Tech, Inc. (Food
resolution by reducing peak width and Drug Administration, Silver Spring, Mary-
Belgium, 2011).
and trailing. land, 2020).
(13) FDA Compliance Program Guide CPG
• Direct resolution enhancement by 7346.832 Pre-Approval Inspections (Food (32) FDA Warning Letter Fresenius Kabi Oncol-
even derivative peak sharpening and Drug Administration: Silver Spring, ogy (Food and Drug Administration: Silver
Maryland, 2019). Spring, Maryland, 2017) Available from:
also increases resolution by reducing https://www.fda.gov/ICECI/EnforcementAc-
peak width. (14) R.D. McDowall, Spectroscopy 34(12), 14–19 tions/WarningLetters/2017/ucm589941.htm.
(2019).
It is beyond the scope of this article (33) R.D.McDowall, LCGC Europe 27(9), 486–492
(15) MHRA GxP Data Integrity Guidance and Defini- (2014).
to present and discuss what is already in tions (Medicines and Healthcare Products Regu-
Wahab’s paper, but if any of these tech- latory Agency, London, United Kingdom, 2018). (34) R.D. McDowall, LCGC North. Am. 38(3), 163–
172, 193 (2020).
niques are integrated into a chromatogra- (16) FDA Guidance for Industry- Data Integrity and
Compliance With Drug CGMP Questions and (35) M.F. Wahab, G. Hellinghausen, and D.W.
phy data system, then their use needs to Armstrong, LCGC Europe 32(s5), 22–28 (2019).
Answers (Food and Drug Administration: Silver
be justified scientifically. This means from Spring, Maryland, 2018).
development through validation to use of (17) WHO Technical Report Series No.996 Annex
a method. R.D. McDowall
5 Guidance on Good Data and Records Man-
is the director of R.D. McDowall Lim-
If regulators are worried by peak inte- agement Practices (World Health Organisation: ited in the UK. Direct correspondence
gration now, they could be paranoid in Geneva, Switzerland, 2016). to: rdmcdowall@btconnect.com
the future! (18) PIC/S PI-041-3 Good Practices for Data Man-
SUPPLEMENT TO
THE
APPLICATION
NOTEBOOK
June 2020
www.chromatographyonline.com
TABLE OF CONTENTS
THE APPLICATION
NOTEBOOK
Environmental
357 The Extraction of PFAS Molecules from Spiked Soil
Alicia D. Stell, CEM Corporation
Medical/Biological
359 Tailored Analysis of Phospholipid Classes Using
iHILIC-Fusion(+) as First Dimension for Online
Two-Dimensional HILIC–Reversed-Phase LC–MS
Patrick O. Helmer*, Carina M. Wienken*, Wen Jiang†, and Heiko Hayen*,
*Institute of Inorganic and Analytical Chemistry, University of Münster, Münster,
Germany, †Hilicon AB
Pharmaceutical/Drug Discovery
366 Amino Acid Analysis by European Pharmacopeia
to Support Coronavirus Research
Pickering Laboratories, Inc.
367 Size-Exclusion Chromatography for the Impurity
Analysis of Adeno-Associated Virus Serotypes
Stephan M. Koza and Weibin Chen,Waters Corporation
369 An Anion-Exchange Chromatography Method
for Monitoring Empty Capsid Content in
Adeno-Associated Virus Serotype AAV8
Hua Yang, Stephan M. Koza, and Weibin Chen,Waters Corporation
Per- and polyfluoroalkyl substances (PFAS) are a group of manufactured Sample Preparation and EDGE Method
chemicals that are used in a wide variety of industries because of their Five grams of clean sandy loam was weighed directly into a Q-Cup®
resistance to stains, grease, and high temperatures. They possess a chain containing the S1 Q-Disc® stack (C9-G1-C9 sandwich). Each sample
of linked carbon atoms with fluorine atoms branching off the main chain. was spiked with 2 ng or 200 ng of standard in HPLC-grade methanol,
The presence of the strong carbon-fluorine bond contributes to the stability resulting in low spike and high spike samples, respectively. Each set of
of these compounds, earning them the nickname “forever chemicals.” spikes was done in triplicate. The sample was extracted with the EDGE
PFAS are used in products such as nonstick cookware, firefighting foam, using 80:20 methanol:water with 0.3% ammonium hydroxide using the
and stain-resistant carpets. method provided. Each set of spikes was extracted on the EDGE, and then
Because of their persistent nature and their widespread use, this group a blank extraction of the system was done with a Q-Cup containing the S1
of substances has leached into the environment with limited methods Q-Disc stack (C9-G1-C9 sandwich) to assess the level of carryover. Each
of remediation. Furthermore, these compounds have been found to extraction was collected in a polypropylene conical tube using an EDGE
bioaccumulate in animals and humans, and exposure in humans has rack. The sample was brought up to a final volume of 20 mL using 80:20
been shown to cause adverse health outcomes, including cancer, infertility, methanol:water with 0.3% ammonium hydroxide, and 20 µL of formic
and endocrine disruption. Thus, the assessment of the levels of PFAS in acid was added to each sample to neutralize the sample. The samples
the environment is important to the health and safety of humans. were then analyzed by Pace Analytical.
The EDGE, an automated solvent extraction system, was used to extract
a subset of PFAS molecules from spiked soil samples. The EDGE was EDGE Method
able to extract the soil samples in less than 10 min. The extraction yielded Q-Disc: S1 stack (C9-G1-C9 sandwich)
excellent recoveries and standard deviations. Furthermore, there was no Extraction Solvent: 80:20 methanol:water with 0.3% ammonium
carryover found between samples. The EDGE is an excellent choice for hydroxide
laboratories seeking to automate their PFAS extraction. Cycle 1
Top Add: 10 mL
Method Bottom Add: 0 mL
Rinse: 0 mL
Reagents Temperature: 65 °C
Clean sandy loam was purchased from MilliporeSigma. A PFAS standard Hold Time: 3 min
containing 24 different compounds (Part number 99207) was purchased Cycle 2
from Absolute Standards, Inc. HPLC-grade methanol, HPLC-grade water, Top Add: 10 mL
HPLC-grade formic acid, and ammonium hydroxide were purchased from Bottom Add: 0 mL
Fisher Scientific. Rinse: 0 mL
Table I: Average recovery results from the low spike and high spike soil samples
1H, 1H, 2H, 2H-perfluorodecane sulfonic acid (8:2 FTS) 83.00% 7.81% 87.00% 7.00%
1H, 1H, 2H, 2H-perfluorooctane sulfonic acid (6:2 FTS) 80.00% 3.00% 87.67% 7.09%
1H,1H,2H,2H-perfluorohexane sulfonic acid (4:2 FTS) 86.67% 10.02% 87.67% 5.13%
N-ethylperfluoro-1-octanesulfonamidoacetic acid (EtFOSAA) 85.67% 6.66% 86.00% 5.57%
N-methylperfluoro-1-octanesulfonamidoacetic acid (MeFOSAA) 78.67% 1.15% 81.33% 4.16%
Perfluoro-1-butanesulfonic acid (PFBS) 78.00% 3.00% 86.00% 1.73%
Perfluoro-1-decanesulfonic acid (PFDS) 77.00% 6.56% 84.33% 0.58%
Perfluoro-1-heptanesulfonic acid (PFHpS) 81.00% 2.65% 87.00% 2.65%
Perfluoro-1-nonanesulfonic acid (PFNS) 76.00% 6.56% 84.67% 2.89%
Perfluoro-1-octanesulfonamide (PFOSA) 81.00% 8.00% 86.67% 5.03%
Perfluoro-1-pentanesulfonic acid (PFPeS) 78.00% 4.36% 86.33% 0.58%
Perfluorohexanesulfonic acid (PFHxS) 98.33% 2.89% 89.67% 3.79%
Perfluoro-n-butanoic acid (PFBA) 85.65% 4.01% 88.17% 2.20%
Perfluoro-n-decanoic acid (PFDA) 101.33% 7.51% 93.33% 2.89%
Perfluoro-n-dodecanoic acid (PFDoA) 79.67% 7.37% 79.33% 2.08%
Perfluoro-n-heptanoic acid (PFHpA) 90.67% 9.02% 81.67% 3.21%
Perfluoro-n-hexanoic acid (PFHxA) 100.67% 9.02% 91.08% 6.64%
Perfluoro-n-nonanoic acid (PFNA) 96.33% 3.21% 92.67% 2.52%
Perfluoro-n-octanoic acid (PFOA) 82.77% 3.87% 85.53% 2.25%
Perfluoro-n-pentanoic acid (PFPeA) 79.50% 0.71% 85.67% 4.04%
Perfluoro-n-tetradecanoic acid (PFTeDA) 86.67% 4.73% 88.67% 1.15%
Perfluoro-n-tridecanoic acid (PFTrDA) 63.00% 3.46% 68.33% 3.51%
Perfluoro-n-u0ecanoic acid (PFUdA) 76.33% 1.53% 79.67% 3.51%
Perfluorooctanesulfonic acid (PFOS) 84.00% 4.00% 79.60% 0.53%
Temperature: 65 °C Conclusion
Hold Time: 4 min The analytical assessment of PFAS compounds is critical because of their
Wash 1 widespread nature, high stability, and adverse health effects. The EDGE
Wash Solvent: Methanol was able to rapidly and efficiently extract spiked soil samples with excel-
Wash Volume: 10 mL lent recoveries and RSD values. The EDGE also saw no carryover after
Temperature: 50 °C extractions into the subsequent extraction. The EDGE is an excellent ex-
Hold: 3 s traction tool for laboratories seeking to automate their PFAS extractions
Wash 2 with great efficiency.
Wash Solvent: 80:20 methanol:water with 0.3% ammonium hydroxide
Wash Volume: 10 mL
Temperature: None
Hold: None
Results
The recovery data in Table I, from the low and high spikes, indicated that
the samples were extracted with high efficiency, with recoveries ranging
from 63% to 101%. The resulting RSD values were also low, indicating
the recovery data were reproducible. The extraction data from the empty CEM Corporation
Q-Cup indicated that there was no carryover of the spiked compounds 3100 Smith Farm Road, Matthews, NC 28104
within the system, indicating the wash was aggressive enough to remove tel. (800) 726-3331
any residual PFAS compounds. Website: www.cem.com
Phospholipids (PLs) are a large lipid subgroup that is involved 1D HILIC Separation:
in important cell functions in all organisms. They are the Column: 20 × 2.1 mm, 5-µm, 100 Å, iHILIC®-Fusion(+) (P/N
main components of biological membranes and are essential 100.022.0510, HILICON)
for many biologic processes within and between cells, for Eluents: A) ammonium formate solution (35 mM, pH 3.5) and
example, signal transduction, cell growth, and various 95:5 (v/v) acetonitrile; B) acetonitrile
transport processes. The basic building blocks for PLs are fatty Gradient Elution: 0–0.2 min, 97% B; 0.2–0.5 min, from 97%
acids linked to a glycerol backbone and polar head groups to 93% B; 0.5–2.75 min, 93% B; 2.75–7.5 min, from 93% to
(Figure 1). In addition to the main membrane PLs, such as 60% B; 7.5–11 min, 60% B; 11–11.5 min, from 60% to 97%
phosphatidylcholine (PC) and phosphatidylethanolamine B; re-equilibration is parallel to reversed-phase LC separation
(PE), other PL classes are of great importance as well. Some at 97% B.
PLs have significantly lower concentrations. For example, Flow Rate: 0.3 mL/min; 0.05 mL/min (during parallel
in eukaryotic organisms, cardiolipin (CL) is exclusively reversed‑phase LC separation)
located in the inner mitochondrial membrane (1). Thus, Column Temperature: 40 °C
the concentration of CL in total lipid extracts is very low in Injection Volume: 2–20 µL
comparison to the major membrane lipids, such as PLs and Heart-Cut Setup for PL Transfer: An online heart-cut 2D
cholesterol, but also triacylglycerols (TGs) (2,3). Furthermore, HILIC–reversed-phase LC–MS setup was developed for PL
ion suppression effects complicate the mass spectrometric
(MS) analysis, and a tailored analysis of low abundant PL
classes is often required.
In our previous study, we demonstrated the advantage
of performing sample preparation by solid-phase extraction
(SPE) in hydrophilic interaction liquid chromatography
PC
(HILIC) mode (4). The nonpolar fraction of lipids was
first separated from polar PLs using an offline HILIC SPE
PE
method, and thereafter separation by reversed-phase liquid
chromatography (LC) was performed. Offline methods often
require solvent evaporation and reconstituting analytes in
a suitable solvent for the second dimension, which is a CL
time‑consuming approach and risks losing labile PLs. In this
application, a novel online heart-cut two-dimensional (2D)
HILIC–reversed-phase LC–MS method is presented for the
separation of PL classes and nonpolar lipids, where HILIC was
polar
utilized in the first dimension (1D) and reversed-phase LC in
second dimension (2D). TG cholesterol
Experimental
Lipid Standards: Triacylglycerol (TG 48:0), cholesterol (Chol),
phosphatidylcholine (PC 32:0), phosphatidylethanolamine
(PE 32:0), cardiolipin (CL 64:4), and yeast total lipid extract
(Saccharomyces cerevisiae).
LC–MS/MS Setup: Thermo Scientific Ultimate 3000 system nonpolar
with dual gradient pump hyphenated to a Q Exactive™ Plus
Hybrid Quadrupole-Orbitrap™ mass spectrometer. Ionization
Figure 1: Selected structures of nonpolar lipids and polar PLs. HILIC
was performed utilizing electrospray ionization in negative separation of PLs takes place according to their polar head groups
ionization mode as described by Helmer et al. (3). (shown in green).
2 1
Position A: 2_1
3 Valve 6
Position B: 1_6
4 5
Pump 2 RP-MS
HILICON AB
Tvistevägen 48, SE-90736 Umeå, Sweden
Tel.: +46 (90) 193469
Figure 2: Valve setup for the transfer of PL classes from 1D HILIC E-mail: info@hilicon.com
to 2D reversed-phase (RP) LC via heart-cut approach. Website: www.hilicon.com
rapid, high patient sample throughput that lends itself to automation, but 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Time (min)
5.5 6.0 6.5 7.0 7.5
are compromised by problems with assay interference and by changes Figure 1: Representative LC–MS chromatogram of thyroid hormones
in protein levels that alter free hormone availability (1). in human serum with RPA online cartridge.
Liquid chromatography–mass spectrometry (LC–MS) can offer superior
specificity and speed over immunoassays for determination of thyroid time. We used RPA online cartridges with LC–MS analysis for the
hormones in biological matrices such as serum and tissues. The present detection of thyroid hormones from human serum. The samples were
work demonstrates successful online solid-phase extraction (SPE) with LC– protein precipitated with methanol containing ammonium formate
MS for rapid determination of T4, T3, and 3,3’,5’-triiodo-L- thyronine (rT3) and after centrifugation then directly injected for online SPE and
from biological matrices. LC–MS analysis. Sample loading–washing was performed entirely by
the instrument, eliminating the time-consuming solvent evaporation
Experimental and reconstitution steps.
Materials: Supel™ Genie RP-Amide (RPA) and C8 (results not shown) The RPA cartridges captured a trace amount (100 ng/mL × 2 µL in
online SPE cartridges (2 cm × 4.0 mm i.d.); human serum; and protein this case) of thyroid hormones from human serum (Figure 1). All three
precipitation solvent: methanol with 1% (w/v) ammonium formate. analytes were well resolved with a peak width at half height <6 s and
Sample Processing: Human serum spiked with analytes was protein tailing factor from 1.5–1.8. The total run time was within 6 min.
precipitated by vortex mixing with the precipitation solvent at a 1:3 An online SPE–LC–MS method can be used for rapid detection of
ratio. The mixture was then centrifuged at 10,000 ×g for 3 min and thyroid hormones in human serum with minimal hands-on effort and
the supernatant was collected and directly injected for LC–MS analysis. time-consuming steps.
Online SPE–LC–MS Setup: Six-port switching valve and two pumps; one
for sample loading and washing, the other for sample elution. To minimize Reference
potential peak broadening from the cartridges, the flow of sample loading– (1) N. Kahric-Janicic, S.J. Soldin, O.P. Soldin, T. West, J. Gu, and J. Jonklaas,
washing and the subsequent elution are in reversed directions. Thyroid 17(4), 303–11 (2007).
Table I: Ruggedness of online SPE–LC–MS with RPA cartridges from 120 consecutive injections of human serum samples.
Peak Area
Retention Time (Min) Retention Time Reproducibility Peak Area
Analyte MRM Quantifier Reproducibility
(Avg. n = 120) (%RSD, n = 120) (Avg. n = 120)
(%RSD, n = 120)
3,3’,5-triiodo-L-thyronine (T3) 651.8 / 605.5 4.03 0.2 27046 5.1
3,3’,5-triiodo-L-thyronine (rT3) 651.8 / 605.5 4.43 0.2 33723 6.2
3,3’,5,5’-tetraiodo-L-thyronine (T4) 777.7 / 731.8 4.79 0.2 23766 7.7
with a Waters Xevo TQ-S mass spectrometer. Instrument conditions to 50 ng/mL for fentanyl, alfentanil, acetyl fentanyl, butyryl fen-
were as follows, and analyte transitions are provided in Table I. tanyl, and sufentanil; from 0.10 to 50 ng/mL for remifentanil; and
from 0.25 to 50 ng/mL for norfentanyl and carfentanil. All analytes
Analytical column: Raptor Biphenyl (5 µm, showed acceptable linearity with r2 values of 0.996 or greater and
50 mm × 2.1 mm; cat. #9309552) deviations of <12% (<20% for the lowest concentrated standard).
Guard column: Raptor Biphenyl EXP guard column
cartridge, (5 µm, 5 mm × 2.1 mm; Accuracy and Precision
cat. #930950252) Based on three independent experiments conducted on multiple
Mobile phase A: 0.1% Formic acid in water days, method accuracy for the analysis of fentanyl and its ana-
Mobile phase B: 0.1% Formic acid in methanol logues was demonstrated by the %recovery values, which were
Gradient Time (min) %B within 10% of the nominal concentration for all compounds at all
0.00 30 QC levels. The %RSD range was 0.5–8.3% and 3.4–8.4% for in-
2.50 70 traday and interday comparisons, respectively, indicating accept-
2.51 30 able method precision (Table II).
3.50 30
Flow rate: 0.4 mL/min Conclusions
Injection volume: 5 µL A simple dilute-and-shoot method was developed for the quantita-
Column temp.: 40 °C tive analysis of fentanyl and its analogues in human urine. The
Ion mode: Positive ESI analytical method was demonstrated to be fast, rugged, and sen-
sitive with acceptable accuracy and precision for urine sample
Results analysis. The Raptor Biphenyl column is well suited for the analy-
Chromatographic Performance sis of these synthetic opioid compounds, and this method can
All eight analytes were well separated within a 2.5-min gradient be applied to clinical toxicology, forensic analysis, workplace drug
elution (3.5-min total analysis time) on a Raptor Biphenyl column testing, and pharmaceutical research.
(Figure 1). No significant matrix interference was observed to neg-
atively affect quantification of the fivefold diluted urine samples.
The 5-µm particle Raptor Biphenyl column used here is a superfi-
cially porous particle (SPP) column. It was selected for this meth-
od in part because it provides similar performance to a smaller
particle size fully porous particle (FPP) column, but it generates
less system back pressure.
Linearity
Linear responses were obtained for all compounds and the calibra- Restek Corporation
tion ranges encompassed typical concentration levels monitored 110 Benner Circle, Bellefonte, PA 16823
for both research and abuse. Using 1/x weighted linear regression tel. 1 (814) 353-1300
(1/x2 for butyryl fentanyl), calibration linearity ranged from 0.05 Website: www.restek.com
Table II: Accuracy and precision results for fentanyl and related compounds in urine QC samples
Mass (g/mol)
DŽůĂƌDĂƐƐ;ŐͬŵŽůͿ
Technology) coupled to size-exclusion chromatography (SEC). A 1.0x104ϰ
ϭ͘ϬdžϭϬ
differential refractive index (dRI) detector served to measure
dRI Signal
ĚZ/^ŝŐŶĂů
concentration. MALS and dRI measurements are combined in
Molecular
ϭ͘ϬdžϭϬ
1.0x103ϯ
ASTRA software to calculate molar mass distributions, moments
such as the weight-average molar mass Mw, and the polydispersity
index Mw/Mn. 1.0x102Ϯ
ϭ͘ϬdžϭϬ
Analysis by MALS relies on first principles and therefore
overcomes the uncertainties associated with column calibration, 1.0x101ϭ
ϭ͘ϬdžϭϬ
which arise from differences in elution properties between the 6.0
ϲ͘Ϭ 8.0
ϴ͘Ϭ 10.0
ϭϬ͘Ϭ 12.0
ϭϮ͘Ϭ 14.0
ϭϰ͘Ϭ
dŝŵĞ;ŵŝŶͿ
Time (min)
sample and reference molecules related to conformation, density,
and column interactions. SEC–MALS also eliminates the need to
perform frequent recalibration. Figure 1: Molar mass versus time for two chondroitin sulfate samples
measured by SEC–MALS. The dRI signals for both samples and the LS signal
for one sample are superimposed.
Results
Figure 1 shows chromatograms and molar masses obtained from
MALS measurements for two samples, CS and CSLM. The dRI trace Conclusions
alone is shown for CSLM in red, while both dRI and MALS signals The SEC–MALS method provides complete information about the mo-
are shown for CS, in blue. Not all of the differences between the two lar mass of chondroitin sulfate, determining distributions and a full
samples appear in standard SEC analysis, but are readily revealed set of moments. The analysis indicated similarities and discrepancies
by SEC–MALS. between the two samples. The presence of a high-molecular-weight
MALS indicates that CS has a Mw of 14.8 ± 0.2 kDa and the peak in CS, not observed by dRI, was detected unequivocally by MALS,
polydispersity index (Mw /Mn) is 1.21 ± 0.03. The low molar mass thanks to its enhanced sensitivity to high-molecular-weight species.
chondroitin sulfate LMCS sample has a Mw of 6.9 ± 0.2 kDa and a The fact that similar molar masses and sizes of each sample were
polydispersity of 1.4 ± 0.1. MALS analysis in fact determines a full eluted at the same time is a good indication of similar conformation
set of moments including Mn, Mw, and Mz, as well as cumulative and and chemical properties. Absolute macromolecular characterization
differential molar mass distributions (not shown). by SEC–MALS provides rich, detailed information without the biases
inherent in standard analytical SEC, for confidence in essential
• Two additional peaks were observed in both samples at late physiochemical properties.
elution times, though at different ratios to the main peak. The
molar masses of these species are calculated by MALS as just
a few hundred g/mol.
• In the CS sample, an early peak at 7 min is observed in the
MALS trace but is barely present in the dRI trace. This high ratio
of MALS to dRI is typical of high-molar-mass species; further
analysis shows that this peak contains a fraction with molar
mass ranging from 40–500 kDa. Wyatt Technology Corporation
6330 Hollister Ave., Santa Barbara, CA 93117
tel. (805) 681-9009
Website: www.wyatt.com
Experimental Conditions
The analysis of glucose-6-phosphate, fructose-6-phosphate, glucose- TM
Ammonia
COVID-19. The focus on biologic drugs during the current
pandemic highlights the crucial role amino acids analysis
plays in research and production of pharmaceuticals.
Lysine
Threonine
The European Pharmacopoeia (Ph. Eur.) defines requirements for
Methionine
Phenylalanine
Isoleucine
Valine
the qualitative and quantitative composition of medicines, as well
Leucine
as the tests to be carried out on medicines and on substances and
materials used in their production.
Proline
Pickering Laboratories, Inc. offers a complete solution for amino
acids analysis according to Ph. Eur.. This includes the Onyx PCX 0 40 min
10 20 30
post-column derivatization instrument, analytical columns and
GARDs, buffers, and Trione® ninhydrin reagent. The methods pre- Figure 2: Sodium chromatogram of alternative amino acids ana-
lyzed using Ph. Eur. 9.0 methods (3 ug/mL each, 50 uL injection).
sented in this application note were optimized to comply with system
suitability requirements of Ph. Eur. 9.0 methods.
Ammonia
Equipment
• Quaternary HPLC pump, autosampler, UV-vis detector
• Onyx PCX post-column derivatization system
Post-column reagent To make it easier to start using Pickering Laboratories methods, we offer
Trione® ninhydrin reagent chemistry kits that include an analytical column, GARD buffers, and
reagents for amino acids analysis. All parts of the kit could be ordered
individually if needed. Please contact Pickering Laboratories if you have
Ammonia
Alanine
Threonine
Lysine
Isoleucine
Leucine
Valine
Arginine
Cystine
Proline
Monitoring the size heterogeneity of AAV-based gene LC system: ACQUITY™ UPLC™ H-Class Bio
therapy therapeutics is potentially important to ensure Sample temp.: 6 ˚C
consistent product quality and efficacy. We demonstrate Column temp.: 25 ˚C
that the levels of both high-molecular weight (HMW) and Injection volume: 3 µL
low-molecular weight (LMW) impurities in AAV capsid Column: XBridge™ Protein BEH SEC, 450 Å,
preparations can be separated on a Waters™ 450 Å Protein 3.5 µm, 7.8 × 300 mm
BEH SEC column for a series of AAV serotypes. Fluorescence
detector: Excitation: 280 nm; Emission: 350 nm
As the development of gene therapy products accelerates, the Mobile phase: 10 mM NaH2PO4, 10 mM Na2HPO4,
need to develop sound and efficient analytical strategies to 200–400 m KCl, pH 6.6 (HCl)
help guide the development of manufacturing processes and Flow rate: 0.6 mL/min
evaluate the quality of clinical adeno-associated virus-(AAV)–
based gene therapy materials has become more important. Results and Discussion
Among other critical quality attributes, the levels of potential The upper analyte size limitation for SEC separations of
AAV high molecular weight (HMW) species and AAV fragments proteins is approximately 100–200 nm (subvisible) depending
or low molecular weight (LMW) species may also require on SEC particle size. Above these limits, the analyte may be
monitoring (1). Here we present optimized non-denaturing altered due to shear forces or trapped by the frits or packed
size-exclusion chromatography (SEC) methods that can bed of the column. Therefore, these subvisible entities are
separate soluble AAV for several AAV-CMV-GFP serotypes analyzed using dynamic light scattering and nanoparticle
including AAV1, AAV2, AAV5, AAV6, AAV8, and AAV9. tracking analysis (NTA) methods, among others (2). With these
considerations, the SEC separation of soluble AAV monomer,
Experimental Conditions and HMW, and LMW was evaluated on a 450 Å pore-size
AAV sample: AAV serotypes ranging from 1.6 × 1012 to BEH diol-bonded SEC column. This column was previously
6.7 × 1013 GC/mL demonstrated to be effective in the separation of IgM pentamer
ACQUITY UPLC Protein BEH SEC, 450 Å, 2.5 µm, 4.6 x 300 mm Column
0.05
Flow rate: 0.35 mL/min 5
0.045 Injection volume: 5 µL
0.04 4
0.035
UV (280 nm)
0.03
0.025 3
0.02
2
0.015 1a
0.01 1b
0.005
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9.5 10 10.5 11 11.5 12 12.5
Minutes
Figure 1: Shown is the SEC separation of several protein structures. Compounds are: 1a. IgM dipentamer (1.8 kDa), 1b. IgM pentamer (900 kDa), 2.
thyroglobulin (667 kDa), 3. apoferritin (443 kDa), 4. β-amylase (200 kDa), 5. IgG (150 kDa). UV absorbance (280 nm) and the identities of the peaks observed
for the IgM sample were confirmed by SEC–MALS analysis.
(g/mol)
provide the accessible pore volume needed. A 3.5 µm particle
Mass (g/mol)
6.0x1066
6.9 E6
6.9 E6 (± 0.3 E6)
E6)
size was selected over 2.5 µm particles to reduce potential
MolarMass
sample sieving and shearing effects. Intrinsic protein
(V)
voltage (V)
6
0.15
0.15
detectorvoltage
4.0x106
Molar
fluorescence detection was also employed to provide maximum 0.10
0.10
3.8
3.8 E6
E6 (±
(± 0.03 E6)
optical sensitivity.
detector
0.05
0.05
6
2.0x106
The SEC separation of a null control sample (AAV without
0.00
0.00
3.0 4.0 5.0 6.0 7.0
7.0 8.0 9.0
Minutes
Minutes
serotypes AAV1, AAV2, AAV5, AAV6, AAV8, and AAV9 (Figure (200 mM KCl) 20.00
0.60
(350 mM KCl)
30.00
20.00
3). We observe that the peak shapes for both the dimer 0.40 10.00
D=1.6%
10.00
0.40
0.00 0.00
0.20 0.20
M=0.5%
F=0.2%
and return to baseline appropriately. In addition, detectable D=0.2%
0.00
0.00
(350 mM KCl)
100.00
0.80 2.00
(400 mM KCl)D=1.5%
D=1.1%
20.00
50.00
0.60
0.20
Conclusions 0.00
0.00
-0.20
AAV9-GFP 100.00
10.00 60.00
D=1.2%
40.00
5.00
1.00
20.00
0.00
M=0.9%
0.00 5.00 10.00 15.00 20.00 25.00
F=0.8%
dimers, lower valency multimers, and LMW fragments. The M=0.3%
F=0.3%
minimal amount of ionic strength (KCl) required for optimal 0.00 0.00
peak shape and recovery varied by serotype, and separations 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00
Minutes Minutes
are presented for serotypes AAV1, AAV2, AAV5, AAV6, AAV8,
and AAV9. Figure 3: Shown are the SEC separations of a series of AAV serotype
control samples containing ssDNA coding for green fluorescent protein
(GFP). The peak percentages for dimer (D), multimer (M), and fragments (F)
References
are provided. The chromatogram baselines are zoomed approximately 50x
(1) J. F. Wright, Gene Therapy 15(11), 840–8 (2008). versus the full-scale chromatogram shown in the inset.
(2) (a) J. F. Carpenter, T. W. Randolph, W. Jiskoot, D. J. Crommelin, C. R.
Middaugh, G. Winter, Y. X. Fan, S. Kirshner, D. Verthelyi, S. Kozlowski, K.
A. Clouse, P. G. Swann, A. Rosenberg, and B. Cherney, J Pharm Sci 98(4),
1201–5 (2009); (b) B. Slutter, and W. Jiskoot, Expert Opin. Drug Delivery
13(2), 167–70 (2016).
(3) R. Saber, S. Sarkar, P. Gill, B. Nazari, and F. Faridani, Sci. Iran. 18(6),
1643–1646 (2011).
(4) S. Vernon, J. Stasny, A. Neurath, and B. Rubin, J. Gen. Virol. 10(3), 267–
272 (1971). Waters Corporation
34 Maple Street, Milford, MA 01757
Waters, The Science of What’s Possible, ACQUITY, UPLC, and XBridge are trademarks tel. 508 478 2000
of Waters Corporation. All other trademarks are the property of their respective owners. Website: www.waters.com
20.00
EU
0.00
20.00
EU
Empty capsid
10.00
0.00
20.00
EU
Full capsid
10.00
0.00
8.00 10.00 12.00 14.00 16.00 18.00 20.00
Minutes
Figure 1: AAV8 ssDNA (CMV-GFP) empty and full capsids are separated on a Protein-Pak Hi Res Q Column using an optimized AEX method (see
experimental conditions for method details).
24.00 100
0% Empty
22.00 90 100 x A Empty
12.5% Empty % Empty
80 RF F/E
Measured % Empty
20.00 25% Empty (A Empty +
18.00 50% Empty 70 A Full
16.00
75% Empty 60
100% Empty 50
14.00
40
EU
12.00
30
10.00
20
8.00 y = 0.9264x + 3.2146
10 R = 0.9997
6.00
0
4.00 0 20 40 60 80 100
2.00 Predicted % Empty
0.00
Figure 2: Shown on the left is an overlay of AEX chromatograms for AAV8 samples with differing levels of empty capsid. On the right is the
correlation between predicted and measured empty capsid content. Calculation of the percentage of empty capsid was based on the equation
shown, which corrects for the difference in fluorescence response between empty and full capsids. AEmpty and AFull are the AEX peak areas of the
empty and full capsids, and RFF/E is the fluorescence response factor.
2.00
4.00
AAV2
2.00
AAV5
of empty capsid levels between samples are required. 0.00
4.00
2.00
AAV8
0.00
Conclusions
AAV9*
EU
5.00
AEX will not likely provide a measurement of AAV8 capsid
0.00
carrying partial ssDNA, therefore complementary methods F
E
10.00
EU
References
(1) X. Fu, et al, Hum. Gene. Ther. Methods 30(4), 144 (2019).
(2) C. Wang, et al, Mol. Ther. Methods Clin. 19, 257 (2019).
(3) H. Yang, et al, Waters Application Note 720006825EN (2020).
Waters Corporation
Waters, The Science of What’s Possible, Protein-Pak, ACQUITY, and UPLC are 34 Maple Street, Milford, MA 01757
trademarks of Waters Corporation. All other trademarks are the property of their tel. 508 478 2000
respective owners. Website: www.waters.com
FUN DA MENTAL S
confirm that sample is being aspirated paying attention to the stationary-phase should also lead a reduction in signal-to-
from the vial and that it the correct volume. film thickness. noise ratio for analyte peaks.
• Check and replace the inlet septum as • Check that the carrier gas is flowing at the
required, and check that the correct liner correct volumetric flow rate using a cali- Tony Taylor is the Chief Sci-
has been installed. brated flow meter. ence Officer of Arch Sciences
Group and the Technical
• Check that the sample preparation or dilu- • Check that the carrier gas flow program-
Director of CHROMacademy.
tions are correct as per the sample prepa- ming method is correct by selecting either Direct correspondence to:
ration method. constant pressure or constant flow oper- LCGCedit@mmhgroup.com.
• With flame based ionizing detectors, ating modes. In the latter, the carrier gas