Analytical and Bioanalytical Chemistry (2022) 414:4391–4399

https://doi.org/10.1007/s00216-021-03785-8

RESEARCH PAPER

Quantitative analysis of short‑chain fatty acids in human plasma

and serum by GC–MS
Linxing Yao1   · Emily A. Davidson2,3 · Maliha W. Shaikh4 · Christopher B. Forsyth4 · Jessica E. Prenni5 ·
Corey D. Broeckling1

Received: 29 September 2021 / Revised: 5 November 2021 / Accepted: 9 November 2021 / Published online: 29 January 2022
© Springer-Verlag GmbH Germany, part of Springer Nature 2021

Abstract
Short-chain fatty acids (SCFAs) are volatile fatty acids produced by gut microbial fermentation of dietary nondigestible
carbohydrates. Acetate, propionate, and butyrate SCFA measures are important to clinical and nutritional studies for their
established roles in promoting healthy immune and gut function. Additionally, circulating SCFAs may influence the metabo-
lism and allied function of additional tissues and organs. The accurate quantification of SCFAs in plasma/serum is critical to
understanding the biological role of SCFAs. The low concentrations of circulating SCFAs and their volatile nature present
challenges for quantitative analysis. Herein, we report a sensitive method for SCFA quantification via extraction with methyl
tert-butyl ether after plasma/serum acidification. The organic extract of SCFAs is injected directly with separation and detec-
tion using a polar GC column coupled to mass spectrometry. The solvent-to-sample ratio, plasma volume, and amount of HCl
needed for SCFA protonation were optimized. Method validation shows good within-day and inter-day repeatability. The
limit of detection was 0.3–0.6 µg/mL for acetate and 0.03–0.12 µg/mL for propionate and butyrate. Successful application
of this method on clinical plasma and serum samples was demonstrated in six datasets. By simplifying the sample prepara-
tion procedure, the present method reduces the risk of contamination, lowers the cost of analysis, increases throughput, and
offers the potential for automated sample preparation.

Keywords  Acetate · Butyrate · Plasma · Propionate · Serum · Short-chain fatty acid

Published in the topical collection Analytical Methods and Applications
in the Materials and Life Sciences with guest editors Ute Resch-Genger,
Matthias Koch, Björn Meermann, and Michael G. Weller.
This manuscript is dedicated to the 150th anniversary of BAM.
* Linxing Yao
Linxing.Yao@colostate.edu
1
Analytical Resources Core–Bioanalysis and Omics,
Colorado State University, Fort Collins, CO 80523, USA
2
Department of Environmental Health Sciences, Yale School
of Public Health, Yale University, New Haven, CT 06510,
USA
3
Department of Cellular & Molecular Physiology, Yale School
of Medicine, Yale University, New Haven, CT 06510, USA
4
Rush University Medical Center and Rush University,
Chicago, IL 60612, USA
5
Department of Horticulture and Landscape Architecture,
Colorado State University, Fort Collins, CO 80523, USA
Introduction
Short-chain fatty acids (SCFAs) are monocarboxylic acids
with a chain length of C2 to C6, mainly derived from gut
microbial fermentation of dietary fiber and nondigestible
polysaccharides [1]. SCFAs, in particular, acetate (C2),
propionate (C3), and butyrate (C4), have been intensely
studied for their health benefits. Known to be involved in
gastrointestinal physiology, SCFAs contribute to gut health
by serving as energy substrate for colonocytes, stimulating
mucus production, promoting mucosal immune and bar-
rier function and collective intestinal homeostasis [2, 3].
Preliminary evidence suggests that butyrate, in particular,
can reduce colorectal cancer risk by inhibiting histone dea-
cetylase activity [4, 5]. A small percentage of gut-derived
SCFAs enter systemic circulation [6]. As summarized in
recent reviews [2, 3, 7], circulating SCFAs have demon-
strated positive effects on the metabolism and allied func-
tion of tissues and organs beyond the colon. SCFAs bind to
several well-characterized metabolite receptors on immune

13
Vol.:(0123456789)
4392
and endocrine cells in tissues and organs throughout the
body and act as important signaling molecules. SCFAs
have shown to reduce the risk of developing type II diabe-
tes, obesity, and other metabolic diseases, although con-
flicting data exist [2, 3, 8, 9]. SCFAs can be transported
across the blood–brain barrier (BBB) into the brain and
cerebrospinal fluid and affect permeability of the BBB and
brain function [2, 7, 10]. Recent research, as reviewed by
Silva et al. [7], suggests that SCFAs act in the development
and homeostasis of the central nervous system (CNS) and
play an important role in gut-brain interactions. SCFAs
may participate in modulation of neurotransmitters and
neurotrophic factors that play important roles in learning,
memory, circadian rhythm, appetite, and brain disorders
[7, 11]. Despite the large number of studies across the clin-
ical and nutrition fields, more research is needed to confirm
the beneficial health effect of SCFAs, especially in human
trials [3]. Further elucidation of the mechanism underlying
SCFAs various positive physiological effects will establish
the foundational knowledge needed to effectively utilize
SCFAs as dietary intervention for disease prevention and
treatment [12]. Consequently, there is a growing demand
for accurate quantification of colonic, fecal, and circulat-
ing SCFAs. Because plasma/serum SCFAs concentrations
are considerably lower than fecal SCFA or colonic SCFA,
an analytical method with high sensitivity and selectivity
is necessary. SCFAs have been analyzed using gas chro-
matography (GC) coupled with either flame ionization
(FID) or mass spectrometry (MS) detectors, and liquid
chromatography coupled with MS (LC–MS). LC–MS
approaches usually involve derivatization to improve ioni-
zation of SCFAs, whereas GC approaches are suitable for
either derivatized or non-derivatized samples [13]. Mass
spectrometry is the preferred detection method owing to its
greater sensitivity and selectivity. Historically, developers
of SCFA quantification methods attempted to popularize
the acidification of the plasma supernatant obtained after
protein precipitation, followed by direct injection of the
extract onto the GC column [14, 15]. Such methods were
abandoned because repeat injections of low pH aqueous
solution led to rapid column deterioration. Subsequently,
the use of organic solvents to extract protonated SCFAs
from acidified plasma/serum became routine. Because
plasma/serum SCFAs are volatile and biologically exist
in low concentrations, additional sample preparation steps
were introduced to improve the assay sensitivity, which
included plasma protein removal [16, 17], neutralization
[17–19], solid-phase extraction [17], and chemical deri-
vatizations [16, 17, 20–25]. Unfortunately, such methods
often require lengthy preparation time, are expensive, and
are accompanied by a high risk of acetate contamination
and increased analytical variation [17, 24]. Therefore, the
aim of the present study was to develop and validate a
13
Yao L. et al.
sensitive and high throughput method allowing for the
rapid extraction of plasma/serum SCFAs to support accu-
rate quantification of SCFAs by GC–MS.
Materials and method
Materials
Human plasma samples (hereafter referred to as control
plasma) used for method development and validation were
obtained from UCHealth Garth Englund Blood Center (Fort
Collins CO, USA) with an approval from the Institutional
Review Board of Colorado State University (IRB#16-063B).
Acetic acid-13C2, acetic acid, propionic acid, butyric acid,
and 5-sulfosalicylic acid were purchased from Sigma-
Aldrich (St. Louis, MO). Sodium butyrate-13C4 was pur-
chased from Santa Cruz Biotechnology (Dallas, TX). Methyl
tert-butyl ether (MTBE), water (Optima UHPLCMS grade),
and hydrochloric acid were purchased from Fisher Scientific
(Hampton, NH).
Plasma sample preparation
Plasma samples stored (in Eppendorf tubes) at − 80 °C were
moved to 4 °C for thawing with intermittent mixing. To the
thawed plasma (200 µL), 24 µL of cold internal standard
solution containing 50 µg/mL of acetic acid-13C2 in 2 N HCl
was added. Samples were gently vortexed for 5 s, followed
by the addition of 360 µL of cold MTBE. The mixture was
vigorously vortexed for 15 s and centrifuged at 3,000 g for
10 min at 4 °C. Approximately, 60 µL of the upper MTBE
layer was recovered and stored at − 20 °C until analysis. This
extraction protocol is referred to as the “standard method”
hereafter and used for all tests unless otherwise specified.
A Quality Control (QC) sample was generated by pooling a
small aliquot of each experimental sample when the sample
set size was small (< 30 samples). For larger sample sets,
control plasma or a subset of samples if available was used
as the QC sample. QC samples were prepared alongside
experimental samples using the standard method.
Calibration curve
Authentic standards of acetic acid, propionic acid, and
butyric acid were diluted in 100% water with concentra-
tions ranging from 0.1 to 100 µg/mL for acetic acid and 0.01
to 25 µg/mL for propionic and butyric acid. Curve aliquots
were extracted with MTBE as described above with the pres-
ence of an internal standard.
Quantitative analysis of short-chain fatty acids in human plasma and serum by GC–MS 4393

Optimization of sample extraction ­ CFA1

Table 1  Data processing method for S
Targets Quantification Confirmation Retention
The amount of HCl needed to acidify plasma, MTBE-to- m/z m/z time (min)
plasma ratio, plasma starting volume, and matrix effect were
Acetic acid 60 45 4.76
examined. The details of the experimental designs can be
Propionic acid 74 45 5.60
found in Electronic Supplementary Material.
Isobutyric acid 88 73 5.89
Butyric acid 60 88 6.53
Method validation
Isovaleric acid 60 87 6.98
Valeric acid 60 73 7.68
Standards were spiked into control plasma to generate low,
Acetic acid-13C2 62 45 4.76
medium, and high levels of spiked samples. Control plasma
Butyric acid-13C4 62 92 5.89
with no spiking is referred to as baseline level. Samples for
method validation were extracted using the standard method 1
 Plasma/serum assay only includes the quantification of acetic, pro-
to evaluate within-day repeatability (six replicates), inter- pionic, and butyric acid, and their signals are normalized to acetic
day repeatability (three different days and four replicates per acid-13C2. The assay for fecal samples includes all six SCFAs, where
signals of C4 and C5 SCFAs were normalized to butyric acid-13C4
day), stability of MTBE extracts stored at − 20 °C (three dif-
ferent days within a week and six replicates per day), instru-
ment stability (eight repeat injections of one sample), and of HCl added and it was determined that approximately
accuracy (four replicates). 24 µmol (as minimum) of HCl was required to maximize
the protonation of SCFAs in 100 µL of plasma (Fig. 1). The
GC–MS analysis extraction efficiency of a liquid–liquid extraction is gener-
ally affected by the ratio of solvent to sample. In the case
The MTBE extract (1 μL) was injected into a Trace 1310 of SCFAs, because these compounds are volatile and the
GC coupled to ISQ-LT MS (ThermoFisher, Waltham, evaporation process needs to be avoided, the peak inten-
MA), at a 5:1 split ratio. The inlet was held at 240  °C. sity was determined by both the solvent-to-sample ratio and
SCFA separation was achieved on a DB-WAXUI column solvent volume. Solvent-to-sample ratio of 10:1 [16], 20:1
(30  m × 0.25  mm × 0.25  µm, Agilent, Santa Clara, CA). [20], 6.7:1 [27], and 1:3 [28] were previously reported in
Oven temperature was held at 100 °C for 0.5 min, ramped to the methods using MTBE or diethyl ether to extract SCFAs
175 °C at a rate of 10 °C/min, and then ramped to 240 °C at from plasma. As shown in Fig. 2, when the ratio of sol-
40 °C/min with a final hold for 3 min. The total run time was vent to sample was greater than 2, the peak intensity started
12.7 min. Helium carrier gas flow was held at 1.2 mL/min. to decrease. Although a higher ratio resulted in a greater
Temperatures of transfer line and ion source were both held extraction efficiency, the concentration of analyte in the
at 250 °C. Single Ion Monitoring (SIM) was used at a rate of organic extract actually decreased due to the dilution effect
10 scans per second under electron impact mode. Quantifica- of a larger solvent volume. For the same reason, multiple
tion and confirmation ions are presented in Table 1. extractions were not considered in this method. Based on
these results, the standard method utilized a ratio of 1.8:1
(or 2:1) to maximize extraction efficiency while maintain-
Results and discussion ing a sufficient practical volume for recovering the upper
organic phase.
Method development The standard method was evaluated on 50, 100, 250, and
500 µL starting volumes of plasma and the results demon-
SCFAs are soluble in solvents of a wide range of polarities strated consistent measured concentrations of SCFAs (data
such as water, alcohols, and MTBE. Because SCFAs are not shown). However, SCFA concentrations measured from
weak acids (with pKa 4.7–4.9), their extraction efficiency the 50-µL extraction tended to have a larger standard devia-
in aqueous-organic liquid extraction is pH dependent. Blood tion than those from higher volumes. In addition, for vol-
has sufficient buffering capacity to maintain a pH of ~ 7.4 umes below 200 µL, extraction and extract recovery required
[26]. Thus, SCFAs are present in plasma primarily as a con- the use of specialty small volume glass vials or glass inserts.
jugated weak base, which is insoluble in MTBE. To increase Additional attempts have been made in our method devel-
SCFA partitioning in MTBE, the pH of plasma needs to opment to remove the neutral lipids in plasma by hexane
be lowered to shift the acid–base equilibrium toward the wash prior to MTBE extraction, or to increase the partition-
weak acid form. Various amounts of HCl were tested to ing of SCFAs in the MTBE layer by adding NaCl. However,
disrupt the buffer system of plasma and convert SCFA to both treatments resulted in plasma gelling (occasionally for
acid forms. SCFA peak intensity increased with the amount hexane treatment), which then made it difficult to recover

13
4394 Yao L. et al.

Fig. 1  Effect of the addition 120

of HCl on the SCFA yield as
measured by signal intensity Acetic acid
(n = 4). The peak areas were
normalized to the high- 100 Propionic acid
est values of treatment with
HCl = 24 µmol Butyric acid

Normalized peak area (%)

80

60

40

20

0
3 6 9 12 24 48
HCl (µmol) added to 100 µL of plasma

Fig. 2  Effect of the ratio of 120

MTBE to plasma on the SCFA Acetic acid
signal intensity (n = 4). The
peak areas were normalized to Propionic acid
those obtained from ratio 2:1 100
Butyric acid
Normalized peak area (%)

80

60

40

20

0
1.6:1 1.8:1 2:1 2.5:1 4:1 8:1
Ratio of MTBE-to-plasma (v/v)

MTBE layer. High-speed centrifugation (> 15,000 g) might
help with phase separation of the gelled plasma, but plastic
tubes should be avoided to minimize contamination.
Background noise derived from environmental contami-
nation proved the most challenging aspect of quantifying
plasma/serum SCFAs. Contamination was greatest for acetic
acid as demonstrated by its highly positive y-intercept of the
calibration fit equation (Fig. S1). Background acetate is a
commonly reported methodological challenge and is, seem-
ingly, laboratory and method specific [16, 17, 23–25, 27].
Possible sources of acetate include air, reagents, glassware,
and plastics used during sample preparation. Ethyl acetate,

13
Quantitative analysis of short-chain fatty acids in human plasma and serum by GC–MS 4395

for example, was used as the extraction solvent in a previous protocol. Our standard method yielded the lowest ratio of
study [29]; however, in our study, we found experimentally acetate in blank to acetate in control plasma when compared
that ethyl acetate could contain a high level of acetic acid to other methods (Fig. 3). While derivatization of SCFAs
depending on the age of the solvent, its purity, and storage yielded increased detection, for example, the peak area of
conditions. To further investigate, we systematically exam- acetate in control plasma detected using ref [16] was five-
ined the level of background acetate in the reagents used in fold higher than our method, this came at the cost of high
our method, i.e., the background acetate level in (i) MTBE background acetate.
(direct injection of MTBE), (ii) water (water extracted with While exploring sources of acetate contamination, addi-
MTBE), (iii) HCl (water acidified with HCl and subse- tional precautions were identified as important to minimize
quently extracted with MTBE), (iv) internal standard (water overall background contamination.
acidified with HCl solution containing internal standard and
subsequently extracted with MTBE) — extraction blank. (1) Due to SCFA’s volatile nature, it was important to avoid
Acetate was detected at very low levels or undetected in the use of acetic acid, ethyl acetate, and sodium acetate
neat MTBE (experiment i). Conversely, acetate was detected in the room during sample preparation for SCFA analy-
in water and water containing reagents (experiments ii, iii, sis.
and iv) (Fig. S2 and ref. [30]). Note that the internal stand- (2) The GC syringe required sequential rinsing with metha-
ard, directly dissolved in MTBE, had low levels of back- nol using three rinse bottles on the GC autosampler,
ground acetate, i.e., m/z 60 was about 0.7% of the peak area which were replaced with clean solvent each morning.
of labeled acetate m/z 62, suggesting that the 13C-labeled (3) Before each analytical batch, injector temperature was
form was not a major source of background acetate. We have increased to 300 °C for 10 min with split ratio of 100
examined the water from unopened LC–MS grade bottles, or higher to remove non-volatile components in the
Milli-Q water, and water of other purity grades; unfortu- plasma extracts that had accumulated in the split vent
nately, none of the water products gave sufficiently low level during acquisitions.
of acetate noise. Using a similar approach, we examined the (4) Prior to analysis of experimental samples, background
reagents used in other methods, such as NaOH to neutralize acetate was monitored with the injections of extraction
SCFAs before solvent evaporation [18], and 5-sulfosalicylic blank to ensure that the system was “clean.” We also
acid used to precipitate plasma proteins [16], both of which observed that multiple injections of extraction blank
demonstrated high background acetate levels (Fig. S2). In helped to stabilize the signal of background acetate.
general, background acetate increased with each additional (5) Sample preparation was carried out in batches of 24
reagent, suggesting that the best approach to reduce con- samples (with QCs) plus one extraction blank to moni-
tamination would be to use the simplest sample preparation tor the background acetate level during the day.

Fig. 3  Comparison of the levels 100

of background acetate and
Peak area ratio of acetate in extraction blank to acetate in

plasma acetate obtained using

different sample preparation
methods. Ref [21]-derivatiza-
80
tion with pentafluorobenzyl
bromide (PFBBr); ref [16],
derivatization with N-methyl-N-
control plasma (%)

tert-butyldimethylsilyltrifluoro-
acetamide (MTBSTFA); ref 60
[17, 18], acid/base adjustment.
The ratio was calculated using
the averaged peak area of three
replicates. The relative standard
40
deviation of raw peak areas of
all treatments was < 19%

20

0
Our method Ref [21] Ref [16] Ref [17,18]

13
4396
1400000
1200000
1000000
800000
Peak area
600000
400000
200000
0 0
0 10 20 30 40 50 60
13C -Acetate, µg/mL
2
Fig. 4  Matrix effect on SCFA extraction. 13C2-acetic acid and 13C4-
butyric acid were spiked into control plasma, and water containing
5  µg/mL of acetic acid and 1  µg/mL of butyric acid, respectively.
The concentration of non-labeled acetic acid and butyric acid in
(6) High purity grade of water was used for preparing the
HCl solution and curve samples.
(7) Glass vials (2 mL) were used instead of Eppendorf
tubes because the plastic tube itself could introduce
background acetate (Fig. S3 and ref [24]). While not
tested directly in this study, it has been reported that the
additives used in blood collection tubes also contrib-
uted to the background signal in SCFA analysis [31].
Matrix effect
Various amounts of standards acetic acid-13C2 and butyric
acid-13C4 were spiked into control plasma, and water con-
taining 5 µg/mL of acetic acid and 1 µg/mL of butyric acid
(similar levels to the native SCFAs in control plasma) to
examine matrix effect. Acetic acid-13C2 and butyric acid-
13
C4 extracted from plasma had slightly higher response than
those extracted from water (Fig. 4), which was confirmed
by replications carried out at different times. It is possible
that salts present in plasma improved the extraction yield of
SCFAs [32].
Dynamic range
The linear dynamic range of SCFAs examined in this study
were 0.1 to 100 µg/mL for acetic acid, 0.01 to 10 µg/mL
for propionic acid, and 0.025 to 25 µg/mL for butyric acid
with R2 = 0.9986, 0.9976, 0.9966, respectively (Fig. S1).
Calibration curves of SCFAs used for fecal sample sets are
presented in Fig. S4. Two calculation methods were used
to determine the limit of detection (LOD), (method a) 3
times the standard deviation of extraction blanks divided
by the slope of calibration curve, and (method b) 3 times
the noise signal intensity of blanks. The higher of the two
13
Yao L. et al.
1600000
1400000
1200000
1000000
Peak area
800000
600000
water
water
400000
plasma
plasma
200000
0 5 10 15 20 25
70 80 90 100
13C -Butyrate, µg/mL
4
water were approximated to the concentrations of native SCFA in the
control plasma. Error bars represent the standard deviations of four
extraction replications
LOD values are reported. The LOD for acetate in plasma
was 0.3–0.6 µg/mL, and LOD for propionate and butyrate
was 0.03–0.12 µg/mL. The limit of quantification (LOQ) was
calculated as 3 times of LOD. The reported LODs of plasma
assay include 0.01 µg/mL (C2), 0.02 µg/mL (C3), 0.03 µg/
mL (C4) in reference [16], and 0.6 µg/mL (C2), 0.07 µg/mL
(C3), and 0.09 µg/mL (C4) in reference [27], both of which
used MTBSTFA derivatization following ether extraction.
In the report using a similar approach of direct injection of
ether extracts, Baldi et al. [28] obtained LODs of 0.6 µg/
mL (C2) and 0.1 µg/mL (C3 and C4), and those LODs were
the calculated standard deviation of y-intercepts of regres-
sion lines. The LOD of acetic acid, in particular, varied with
the background acetate level at the time when the analysis
was performed. Decision for whether to report a value for
acetate in samples with low acetate concentration was based
on the stability of the background acetate across acquisition
of the sample set. To perform such an evaluation, at least
one extraction blank (with internal standard) was prepared
and injected with every 24 samples (the level of background
acetate was quite stable during acquisition of a small set
with less than 24 samples). Multiple extraction blanks were
also used in calculation of standard deviation of background
acetate for LOD determination.
Method validation
The within-day and inter-day repeatability, instrument
repeatability of consecutive injections, and stability of
MTBE extracts stored at − 20 °C were evaluated by running
plasma samples with four concentration levels of SCFAs.
Overall, the CVs were generally below 10% (Table  2).
Although the use of Eppendorf tubes increased the level of
Quantitative analysis of short-chain fatty acids in human plasma and serum by GC–MS 4397

Table 2  Assay ­performance1
Plasma spiking level Acetic acid Propionic acid Butyric acid

Within-day repeatability Baseline 3.4 4.1 6.7

(n = 6), Coefficient of variance (CV, %) Low 2.4 3.9 5.6
medium 4.6 6.3 8.7
High 4.7 3.9 3.7
Inter-day repeatability Baseline 7.4 7.3 12.0
(day 0, 3, 7, n = 4 per day), CV (%) Low 7.4 9.4 10.5
medium 4.9 6.3 7.4
High 4.0 4.9 5.4
Stability of MTBE extracts stored at − 20 °C Baseline 4.3 5.5 9.1
(day 0, 3, 7, n = 6 per day), CV (%) Low 3.7 5.2 7.2
medium 5.0 7.4 8.4
High 5.0 5.7 5.2
Instrument repeatability Baseline 3.1 6.7 7.2
(n = 8, repeated injections), CV (%) Low 6.7 2.2 5.8
medium 1.5 2.4 4.2
High 0.3 6.0 1.0
Accuracy (error %) Low 8.8  − 13.9  − 10.2
(n = 4) medium 4.6  − 10.3  − 8.6
High 1.7  − 10.9  − 10.6
1
 Baseline level (control plasma) contained acetic acid − 7.6 µg/mL, propionic acid − 0.1 µg/mL, and butyric acid − 0.2 µg/mL (averaged values
with n = 4). Low level contained acetic acid − 12.5 µg/mL, propionic acid − 0.6 µg/mL, and butyric acid − 0.7 µg/mL. Medium level contained
acetic acid − 27.2 µg/mL, propionic acid − 2.1 µg/mL, and butyric acid − 2.1 µg/mL. High level contained acetic acid − 56.6 µg/mL, propionic
acid − 5.0 µg/mL, and butyric acid − 5.1 µg/mL. Accuracy (error%) = (measured value − theoretical value)/theoretical value

background acetate, the CV of the repeatability test carried
out in plastic tubes were below 7% [30]. Accuracy, expressed
as error %, was calculated as the difference of measured
value and theoretical value divided by theoretical value.
Acetate has 2–9% error, whereas propionate and butyrate
acid have slightly higher error at 9–14%. Accuracy of pro-
pionic and butyric acid may be improved by including pro-
pionic acid-13C3 and butyric acid-13C4 as the internal stand-
ards. The choice to include more than one internal standard
will vary based on the needs and goals of the research ques-
tion. For quantification of six SCFAs in fecal samples, for
example, we used both acetic acid-13C2 and butyric acid-13C4
as internal standards with butyric acid-13C4 being used for
longer chain SCFAs.
Quality control
Ideally, QC samples are pooled from each study sample,
then processed and analyzed along with study samples,
or less ideally, QC samples are pooled from the ready-
to-inject solvent extracts of each study sample. The first
approach can be applied to small sample sets (i.e., < 30
samples). However, for larger sample sets, it is not feasi-
ble to pool from all study samples due to the risks posed
by active enzyme in thawed samples during lengthy QC
generation. The second approach is also difficult to execute
for SCFA assay because of the high volatility of the MTBE
solvent and SCFAs. Instead, we used the control plasma or
commercially purchased samples as the QC. The standard
method presented here has been used to analyze multiple
plasma/serum sample sets in support of various research
projects (Table 3). The CVs of all three SCFAs in the QC
samples run alongside experimental samples were below
24%. For a large sample set, samples were randomized and
split into multiple analytical batches to allow instrument
cleaning between batches. In the set of 377 samples, for
example, the CVs for acetate and butyrate of QC samples
were below 15% in each individual analytical batch, and
the CV of propionate was 10–26% (Table S1). The stand-
ard method for plasma/serum SCFA analysis was further
adapted to the measurement of other sample types such as
fecal samples and fermentation products (detailed in Sup-
plementary Information). The concentrations of SCFAs in
fecal samples are much higher than those in plasma/serum;
thus, sensitivity and contamination concerns are reduced.
However, fecal sample homogeneity and subsampling bias
remain as concerns and can negatively impact the repro-
ducibility of measurements [33]. The CVs of QC samples
for SCFA analysis in fecal samples were generally below
10% (Table S2).

13
4398 Yao L. et al.

Table 3  CV of QC acquired with experimental human plasma/serum sample ­sets1

Number of Number Sample type CV (%) of QC Concentration (µg/mL) of experimental samples

samples of QC
Acetic acid Propionic acid Butyric acid Acetic acid Propionic acid Butyric acid

377 30 Plasma 12.4 16.5 14.3 LOD–20 LOD–0.2 LOD–0.5

83 7 Plasma 6.5 8.9 12.9 1–6 LOD–0.4 LOD–0.4
80 12 Serum 19.0 23.0 17.4 LOD–62 LOD–0.7 LOD–0.3
207 47 Plasma 17.4 22.2 8.2 4–20 0.2–1.6 LOD–0.4
14 4 Plasma 2.6 2.6 7.5 1–3 LOD–0.2 LOD–0.2
192 43 Plasma 3.4 16.3 23.6 3–45 LOD–1.4 LOD–1.1
1
 Since the cohort of 83 samples had low concentration of acetate, instead of using control plasma, we generated a QC using a subset of experi-
mental samples, and the QC had acetate concentration of 1.8 ± 0.1 µg/mL

Conclusions References

The method presented here for the quantification of plasma/
serum SCFAs by GC–MS was optimized and validated.
Additionally, we share challenges and offer recommenda-
tions for minimizing acetate contamination. Successful
application of this method on clinical samples including
plasma, serum, and feces has been demonstrated. Impor-
tantly, by simplifying the sample preparation procedure
(compared to previous methods), we demonstrate a reduction
in the risk of contamination. This simplified approach lowers
the cost of analysis, increases throughput, and has potential
for being adapted to an automated sample preparation.
Supplementary Information  The online version contains supplemen-
tary material available at https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 00216-0​ 21-0​ 3785-8.
Acknowledgements  We thank the UCHealth Garth Englund Blood
Center (Fort Collins, CO, USA) for providing control plasma samples,
and Jane Andales for suggestions of quality assurance for this assay.
Author contribution  L. Yao and C.D. Broeckling designed experi-
ments, L. Yao and E.A. Davidson conducted the experiments and
acquired the data, L. Yao conducted data analysis, L. Yao, E.A. David-
son, M.W. Shaikh, C.B. Forsyth, J. E. Prenni, C.D. Broeckling contrib-
uted to manuscript writing.
Availability of data and material  All data generated or analyzed during
this study are included in this published article.
Code availability  Not applicable.
10. Stilling RM, van de Wouw M, Clarke G, Stanton C, Dinan
Declarations  and butter of the microbiota-gut-brain axis? Neurochem Int.
Ethics approval  Plasma obtained from local blood bank has an
approval from the Institutional Review Board of Colorado State Uni-
versity (IRB#16-063B).
Conflict of interest  The authors declare no competing interests.
1. Koh A, De Vadder F, Kovatcheva-Datchary P, Bäckhed F. From
dietary fiber to host physiology: short-chain fatty acids as key
bacterial metabolites. Cell. 2016;165:1332–45. https://d​ oi.o​ rg/1​ 0.​
1016/j.​cell.​2016.​05.​041.
2. van der Hee B, Wells JM. Microbial regulation of host physiology
by short-chain fatty acids. Trends Microbiol. 2021;29:700–12.
3. Blaak EE, Canfora EE, Theis S, Frost G, Groen AK, Mithieux G,
Nauta A, Scott K, Stahl B, van Harsselaar J, van Tol R, Vaughan
EE, Verbeke K. Short chain fatty acids in human gut and meta-
bolic health. Benef Microbes. 2020;11:411–55. https://d​ oi.o​ rg/1​ 0.​
3920/​BM2020.​0057.
4. Riggs MG, Whittaker RG, Neumann JR, Ingram VM. n-Butyrate
causes histone modification in HeLa and Friend erythroleukaemia
cells. Nature. 1977;268:462–4. https://d​ oi.o​ rg/1​ 0.1​ 038/2​ 68462​ a0.
5. O’Keefe SJD. Diet, microorganisms and their metabolites, and
colon cancer. Nat Rev Gastroenterol Hepatol. 2016;13:691–706.
https://​doi.​org/​10.​1038/​nrgas​tro.​2016.​165.
6. Cummings JH, Pomare EW, Branch WJ, Naylor CP, Macfar-
lane GT. Short chain fatty acids in human large intestine, portal,
hepatic and venous blood. Gut. 1987;28:1221–7. https://​doi.​org/​
10.​1136/​gut.​28.​10.​1221.
7. Silva YP, Bernardi A, Frozza RL. The role of short-chain fatty
acids from gut microbiota in gut-brain communication. Front
Endocrinol. 2020; 11(25). https://​doi.​org/​10.​3389/​fendo.​2020.​
00025.
8. Sowah SA, Hirche F, Milanese A, Johnson TS, Grafetstätter M,
Schübel R, Kirsten R, Ulrich CM, Kaaks R, Zeller G, Kuhn T,
Stangl GI. Changes in plasma short-chain fatty acid levels after
dietary weight loss among overweight and obese adults over 50
weeks. Nutrients. 2020; 12. https://​doi.​org/​10.​3390/​nu120​20452.
9. Shin C, Lim Y, Lim H, Ahn T-B. Plasma short-chain fatty acids in
patients with Parkinson’s disease. Mov Disord. 2020;35:1021–7.
https://​doi.​org/​10.​1002/​mds.​28016.
TG, Cryan JF. The neuropharmacology of butyrate: the bread
2016;99:110–32. https://​doi.​org/​10.​1016/j.​neuint.​2016.​06.​011.
11. Cryan JF, O’Riordan KJ, Cowan CSM, Sandhu KV, Bastiaanssen
TFS, Boehme M, Codagnone MG, Cussotto S, Fulling C, Gol-
ubeva AV, Guzzetta KE, Jaggar M, Long-Smith CM, Lyte JM,
Martin JA, Molinero-Perez A, Moloney G, Morelli E, Morillas E,
O’Connor R, Cruz-Pereira JS, Peterson VL, Rea K, Ritz NL, Sher-
win E, Spichak S, Teichman EM, van de Wouw M, Ventura-Silva

13
Quantitative analysis of short-chain fatty acids in human plasma and serum by GC–MS
AP, Wallace-Fitzsimons SE, Hyland N, Clarke G, Dinan TG. The
microbiota-gut-brain axis. Physiol Rev. 2019;99:1877–2013.
https://​doi.​org/​10.​1152/​physr​ev.​00018.​2018.
12. Dinan TG, Cryan JF. Microbes, immunity, and behavior: psycho-
neuroimmunology meets the microbiome. Neuropsychopharma-
cology. 2017;42:178–92. https://​doi.​org/​10.​1038/​npp.​2016.​103.
13. Chen Z, Wu Y, Shrestha R. Determination of total, free and esteri-
fied short-chain fatty acid in human serum by liquid chromatog-
raphy-mass spectrometry. Ann Clin Biochem. 2019;56:190–7.
https://​doi.​org/​10.​1177/​00045​63218​801393.
14. Remesy C, Demigne C. Determination of volatile fatty acids in
plasma after ethanolic extraction. The Biochem J. 1974;141:85–
91. https://​doi.​org/​10.​1042/​bj141​0085.
15. Buchanan DN, Thoene JG. Volatile carboxylic acid profiling in
physiological fluids: direct injection into a gas chromatograph/
mass spectrometer. Clin Chim Acta. 1985;145:183–91. https://​
doi.​org/​10.​1016/​0009-​8981(85)​90285-2.
16. Moreau NM, Goupry SM, Antignac JP, Monteau FJ, Le Bizec
BJ, Champ MM, Martin LJ, Dumon HJ. Simultaneous measure-
ment of plasma concentrations and 13C-enrichment of short-chain
fatty acids, lactic acid and ketone bodies by gas chromatography
coupled to mass spectrometry. J Chromatogr B Analyt Technol
Biomed Life Sci. 2003;784:395–403. https://​doi.​org/​10.​1016/​
S1570-​0232(02)​00827-9.
17. Morrison DJ, Cooper K, Waldron S, Slater C, Weaver LT, Preston
T. A streamlined approach to the analysis of volatile fatty acids
and its application to the measurement of whole-body flux. Rapid
Commun Mass Spectrom. 2004;18:2593–600. https://​doi.​org/​10.​
1002/​rcm.​1662.
18. Clausen MR, Mortensen PB, Bendtsen F. Serum levels of short-
chain fatty acids in cirrhosis and hepatic coma. Hepatology.
1991;14:1040–5.
19. McArthur B, Sarnaik AP. Quantification of short-chain fatty acids
in plasma. Clin Chem. 1982;28:1983–4.
20. den Besten G, Lange K, Havinga R, van Dijk TH, Gerding A, van
Eunen K, Müller M, Groen AK, Hooiveld GJ, Bakker BM, Rei-
jngoud D-J. Gut-derived short-chain fatty acids are vividly assimi-
lated into host carbohydrates and lipids. Am J Physiol Gastrointest
Liver Physiol. 2013;305:G900-910. https://​doi.​org/​10.​1152/a​ jpgi.​
00265.​2013.
21. Kage S, Kudo K, Ikeda H, Ikeda N. Simultaneous determination
of formate and acetate in whole blood and urine from humans
using gas chromatography-mass spectrometry. J Chromatogr B
Analyt Technol Biomed Life Sci. 2004;805:113–7. https://d​ oi.o​ rg/​
10.​1016/j.​jchro​mb.​2004.​02.​029.
22. Nayananjalie WAD, Wiles TR, Gerrard DE, McCann MA, Hani-
gan MD. Acetate and glucose incorporation into subcutaneous,
intramuscular, and visceral fat of finishing steers1. J Anim Sci.
2015;93:2451–9. https://​doi.​org/​10.​2527/​jas.​2014−8374.
23. Tomcik K, Ibarra RA, Sadhukhan S, Han Y, Tochtrop GP, Zhang
G-F. Isotopomer enrichment assay for very short chain fatty acids
and its metabolic applications. Anal Biochem. 2011;410:110–7.
https://​doi.​org/​10.​1016/j.​ab.​2010.​11.​030.
4399
24. Tumanov S, Bulusu V, Gottlieb E, Kamphorst JJ. A rapid method
for quantifying free and bound acetate based on alkylation and
GC-MS analysis. Cancer Metab. 2016;4:17. https://​doi.​org/​10.​
1186/​s40170-​016-​0157-5.
25. Zheng X, Qiu Y, Zhong W, Baxter S, Su M, Li Q, Xie G, Ore
BM, Qiao S, Spencer MD, Zeisel SH, Zhou Z, Zhao A, Jia W.
A targeted metabolomic protocol for short-chain fatty acids and
branched-chain amino acids. Metabolomics. 2013;9:818–27.
https://​doi.​org/​10.​1007/​s11306-​013-​0500-6.
26. Gilbert DL. Buffering of blood plasma. Yale J Biol Med.
1960;32:378–89.
27. Pouteau E, Meirim I, Métairon S, Fay L-B. Acetate, propionate
and butyrate in plasma: determination of the concentration and
isotopic enrichment by gas chromatography/mass spectrometry
with positive chemical ionization. J Mass Spectrom. 2001;36:798–
805. https://​doi.​org/​10.​1002/​jms.​181.
28. Baldi S, Menicatti M, Nannini G, Niccolai E, Russo E, Ricci F,
Pallecchi M, Romano F, Pedone M, Poli G, Renzi D, Taddei A,
Calabrò AS, Stingo FC, Bartolucci G, Amedei A. Free fatty acids
signature in human intestinal disorders: significant association
between butyric acid and celiac disease. Nutrients. 2021;13:742.
https://​doi.​org/​10.​3390/​nu130​30742.
29. García-Villalba R, Giménez-Bastida JA, García-Conesa MT,
Tomás-Barberán FA, Espín JC, Larrosa M. Alternative method
for gas chromatography-mass spectrometry analysis of short-chain
fatty acids in faecal samples. J Sep Sci. 2012;35:1906–13. https://​
doi.​org/​10.​1002/​jssc.​20110​1121.
30. Yao L, Davidson EA, Broeckling CD, Prenni JE. A novel GC-MS
assay for quantitation of short chain fatty acids in human plasma.
Citation number 288261, Proceedings of the 65th ASMS Confer-
ence on Mass Spectrometry and Allied Topics, Indianapolis, IN,
June 4–8, 2017.
31. Deroover L, Boets E, Yie Y, Vandermeulen G, Verbeke K. Quan-
tification of plasma or serum short-chain fatty acids: choosing the
correct blood tube. J Nutr Health Food Sci. 2017; 5 1−6. https://​
doi.​org/​10.​15226/​jnhfs.​2017.​001112.
32. Poole CF. Milestones in the development of liquid-phase extrac-
tion techniques, Ed. Poole CF, In Handbooks in Separation Sci-
ence, Liquid-Phase Extraction. Elsevier, 2020, Pages 12–13.
https://​doi.​org/​10.​1016/​B978-0-​12-​816911-​7.​00001-3.
33. Hsu Y-L, Chen C-C, Lin Y-T, Wu W-K, Chang L-C, Lai C-H, Wu
M-S, Kuo C-H. Evaluation and optimization of sample handling
methods for quantification of short-chain fatty acids in human
fecal samples by GC−MS. J Proteome Res. 2019;18:1948–57.
https://​doi.​org/​10.​1021/​acs.​jprot​eome.​8b005​36.
Publisher's Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
13