PROTEIN MEMORY

INDEX

• OVERVIEW

• INTRODUCTION

• PROBLEM

• PROTEIN-BASED MEMORY

• DEVELOPMENT IN THE FIELD OF MOLECULAR

ELECTRONICS

• WHY BACTERIORHODOPSIN?

• BACTERIORHODOPSIN OPTICAL MEMORY

• STRUCTURE OF BACTERIORHODOPSIN

• THE PHOTOCYCLE OF BACTERIORHODOPSIN

• 3-D OPTICAL MEMORIES

• DATA READING TECHNIQUE

• DATA WRITING TECHNIQUE

• DATA ERASING

• DATA REFRESHING

• MERITS

• CONCLUSION

Dept. of ECE 1
 PROTEIN MEMORY

OVERVIEW

While magnetic storage devices have been in use since the middle 1950's, today's
computers and volumes of information require increasingly more efficient and faster
methods of storing data. While the speed of integrated circuit has increased steadily over
the past ten to fifteen years, the limits of these systems are rapidly approaching. In
response to the rapidly changing face of computing and demand for physically smaller,
greater capacity, a number of alternative methods to integrated circuit information storage
have surfaced recently. Protein-based optical memory storage using protein
bacteriorhodopsin is a promising one amongst them. Bacteriorhodopsin is a light-
harvesting protein from bacteria that live in salt marshes that has shown some promise as
feasible optical data storage.

Dept. of ECE 2
 PROTEIN MEMORY

Introduction

Since the dawn of time, man has tried to record important events and techniques
for everyday life. At first, it was sufficient to paint on the family cave wall how one
hunted. Then came people who invented spoken languages and the need arose to record
what one was saying without hearing it firsthand. Therefore, years later, earlier scholars
invented writing to convey what was being said. Pictures gave way to letters which
represented spoken sounds. Eventually clay tablets gave way to parchment, which gave
way to paper. Paper was, and still is, the main way people convey information.

However, in the mid twentieth century computers began to come into general use.
Computers have gone through their own evolution in storage media. In the forties, fifties,
and sixties, everyone who took a computer course used punched cards to give the
computer information and store data. In 1956, researchers at IBM developed the first disk
storage system. This was called RAMAC (Random Access Method of Accounting and
Control).

Since the days of punch cards, computer manufacturers have strived to squeeze
more data into smaller spaces. That mission has produced both competing and
complementary data storage technology including electronic circuits, magnetic media like
hard disks and tape, and optical media such as compact disks.

Today, companies constantly push the limits of these technologies to

improve their speed, reliability, and throughput -- all while reducing cost. The fastest and
most expensive storage technology today is based on electronic storage in a circuit such
as a solid state "disk drive" or flash RAM. This technology is getting faster and is able to
store more information thanks to improved circuit manufacturing techniques that shrink
the sizes of the chip features. Plans are underway for putting up to a gigabyte of data onto
a single chip.

Magnetic storage technologies used for most computer hard disks are the most
common and provide the best value for fast access to a large storage space. At the low
end, disk drives cost as little as 25 cents per and provide access time to data in ten
milliseconds. Drives can be ganged to improve reliability or throughput in a Redundant

Dept. of ECE 3
 PROTEIN MEMORY

Array of Inexpensive Disks (RAID). Magnetic tape is somewhat slower than disk, but it is
significantly cheaper per megabyte. At the high end, manufacturers are starting to ship
tapes that hold 40 gigabytes of data. These can be arrayed together into a Redundant
Array of Inexpensive Tapes (RAIT), if the throughput needs to be increased beyond the
capability of one drive. For randomly accessible removable storage; manufacturers are
beginning to ship low-cost cartridges that combine the speed and random access of a hard
drive with the low cost of tape. These drives can store from 100 megabytes to more than
one gigabyte per cartridge.

Standard compact disks are also gaining a reputation as an incredibly cheap way
of delivering data to desktops. They are the cheapest distribution medium around when
purchased in large quantities ($1 per 650 megabyte disk). This explains why so much
software is sold on CD-ROM today. With desktop CD-ROM recorders, individuals are
able to publish their own CD-ROMs.

With existing methods fast approaching their limits, it is no wonder that a number
of new storage technologies are developing. Currently, researches are looking at protein-
based memory to compete with the speed of electronic memory, the reliability of
magnetic hard-disks, and the capacities of optical/magnetic storage. We contend that
three-dimensional optical memory devices made from bacteriorhodopsin utilizing the two
photon read and write-method is such a technology with which the future of memory lies.

Problem

The demands made upon computers and computing devices are increasing each
year, speeds are increasing at an extremely fast clip. However, the RAM used in most
computers is the same type of memory used several years ago. The limits of making RAM
denser are being reached. Surprisingly, these limits may be economical rather than
physical. A decrease by a factor of two in size will increase the cost of manufacturing of
semiconductor pieces by a factor of 5.

Currently, RAM is available in modules called DIMMs and SIMMs .These

modules can be bought in various capacities from a few hundred kilobytes of RAM to
about 64 megabytes. Anything more is both expensive and rare. These modules are

Dept. of ECE 4
 PROTEIN MEMORY

generally 70ns; however 60ns and 100ns modules are available. The lower the
nanosecond rating, the more the module will cost. Currently, a 64MB DIMM costs over
$400. All DIMMs are 12cm by 3cm by 1cm or about 36 cubic centimeters. Whereas a 5
cubic centimeter block of bacteriorhodopsin studded polymer could theoretically store
512 gigabytes of information. When this comparison is made, the advantage becomes
quite clear. Also, these bacteriorhodopsin modules could also theoretically run 1000 times
faster.

In response to the demand for faster, more compact, and more affordable memory
storage devices, several alternatives have appeared in recent years. Among the most
promising approaches include memory storage using protein-based memory.

Protein-Based Memory

There have been many methods and proteins researched for use in computer
applications in recent years. However, among the most promising approaches, the main
one is 3-Dimensional Optical RAM storage using the light sensitive protein
bacteriorhodopsin.

Bacteriorhodopsin is a protein found in the purple membranes of several species

of bacteria, most notably Halobacterium halobium. This particular bacterium lives in salt
marshes. Salt marshes have very high salinity and temperatures can reach 140 degrees
Fahrenheit. Unlike most proteins, bacteriorhodopsin does not break down at these high
temperatures.

Early research in the field of protein-based memories yielded some serious

problems with using proteins for practical computer applications. Among the most serious
of the problems was the instability and unreliable nature of proteins, which are subject to
thermal and photochemical degradation, making room-temperature or higher-temperature
use impossible. Largely through trial and error, and thanks in part to nature's own natural
selection process, scientists stumbled upon a light-harvesting protein that has certain
properties which make it a prime candidate for computer applications. While
bacteriorhodopsin can be used in any number of schemes to store memory, we will focus
our attention on the use of bacteriorhodopsin in 3-Dimensional Optical Memories.

Dept. of ECE 5
 PROTEIN MEMORY

Development In The Field Of Molecular Electronics

There is a revolution fomenting in the semiconductor industry. It may take 30

years or more to reach perfection, but when it does the advance may be so great that
today's computers will be little more than calculators compared to what will come after.
The revolution is called molecular electronics, and its goal is to depose silicon as king of
the computer chip and put carbon in its place. The perpetrators are a few clever chemists
trying to use pigment, proteins, polymers, and other organic molecules to carry out the
same task that microscopic patterns of silicon and metal do now.

For years these researchers worked in secret, mainly at their blackboards, plotting
and planning. Now they are beginning to conduct small forays in the laboratory, and their
few successes to date lead them to believe they were on the right track. "We have a long
way to go before carbon-based electronics replace silicon-based electronics, but we can
see now that we hope to revolutionize computer design and performance," said Robert R.
Birge, a professor of chemistry, Carnegie-Mellon University, Pittsburgh. "Now it's only a
matter of time, hard work, and some luck before molecular electronics start having a
noticeable impact." Molecular electronics is so named because it uses molecules to act as
the "wires" and "switches" of computer chips. Wires may someday be replaced by
polymers that conduct electricity, such as polyacetylene and polyphenylenesulphide.
Another candidate might be organometallic compounds such as porphyrins and
pthalocyanines which also conduct electricity. When crystallized, these flat molecules
stack like pancakes, and metal ions in their centers line up with one another to form a one-
dimensional wire.

Many organic molecules can exist in two distinct stable states that differ in some
measurable property and are interconvertable. These could be switches of molecular
electronics. For example, bacteriorhodopsin, a bacterial pigment, exists in two optical
states: one state absorbs green light, the other orange. Shinning green light on the green-
absorbing state converts it into the orange state and vice versa. Birge and his coworkers
have developed high density memory drives using bacteriorhodopsin. Although the idea
of using organic molecules may seem far-fetched, it happens every day throughout nature.
"Electron transport in photosynthesis one of the most important energy generating
systems in nature, is a real-world example of what we're trying to do," said Phil Seiden,

Dept. of ECE 6
 PROTEIN MEMORY

manager of molecular science, IBM, Yorkstown Heights, N.Y. Birge, who heads the
Center for Molecular Electronics at Carnegie-Mellon, said two factors are driving this
developing revolution, more speed and less space.

"Semiconductor chip designers are always trying to cram more electronic

components into a smaller space, mostly to make computers faster," he said. "And they've
been quite good at it so far, but they are going to run into trouble quite soon." A few years
ago, for example, engineers at IBM made history last year when they built a memory chip
with enough transistors to store a million bytes if information, the megabyte. It came as
no big surprise. Nor did it when they came out with a 16-megabyte chip.

Chip designers have been cramming more transistors into less space since Jack
Kilby at Texas Instruments and Robert Noyce at Fairchild Semiconductor first showed
how to put multitudes on electronic components on a slab of silicon. But 16 megabytes
may be near the end of the road. As bits get smaller and loser together, "crosstalk"
between them tends to degrade their performance. If the components were pushed any
closer they would short circuit. Physical limits have triumphed over engineering. That is
when chemistry will have its day. Carbon, the element common to all forms of life, will
become the element of computers too. "That is when we see electronics based on
inorganic semiconductors, namely silicon and gallium arsenide, giving way to electronics
based on organic compounds," said Scott E. Rickert, associate professor of
macromolecular science, Case Western Reserve University, Cleveland, and head of the
school's Polymer Micro device Laboratory. "As a result," added Rickert, "we could see
memory chips store billions of bytes of information and computers that are thousands
times faster. The science of molecular electronics could revolutionize computer design."
But even if it does not, the research will surely have a major impact on organic chemistry.
"Molecular electronics presents very challenging intellectual problems on organic
chemistry, and when people work on challenging problems they often come up with
remarkable, interesting solutions," said Jonathan S. Lindsey, assistant professor of
chemistry, Carnegie-Mellon University. "Even if the whole field falls through, we'll still
have learned a remarkable amount more about organic compounds and their physical
interactions than we know now. That's why I don't have any qualms about pursuing this
research." Moreover, many believe that industries will benefit regardless of whether an
organic-based computer chip is ever built. For example, Lindsey is developing an

Dept. of ECE 7
 PROTEIN MEMORY

automated system, as well as the chemistry to go along with it, for synthesizing complex
organic compounds analogous to the systems now available for peptide and nucleotide
synthesis. And Rickert is using technology he developed foe molecular electronic
applications to make gas sensors that are both a thousand times faster and more sensitive
than conventional sensors.

For now, the molecular electronics revolution is in the formative stage, and most
of the investigations are still basic more than applied. One problem with which
researchers are beginning to come to grips, though, is determining the kinds if molecules
needed to make the transistors and other electronic components that will go into the
molecular electronic devices, some of the molecules are like bacteriorhodopsin in that
their two states flip back and forth when exposed to wavelengths of light. These
molecules would be the equivalent of an optical switch on which on state is on and the
other state is off. Optical switches have been difficult to make from standard
semiconductors. Bacteriorhodopsin is the light-harvested pigment of purple bacteria
living in salt marshes outside San Francisco. The compound consists of a pigment core
surrounded by a protein that stabilizes the pigment. Birge has capitalized on the clear cut
distinction between the two states of bacteriorhodopsin to make readable-write able
optical memory devices. Laser disks, are read-only optical memory devices, once encoded
the data cannot be changed. Birge has been able to form a thin film of bacteriorhodopsin
on quartz plates that can then be used as optical memory disks. The film consists of a
thousand one-molecule thick layers deposited one layer at a time using the Langmuir-
Blodgett technique. A quartz plate is dipped into water whose surface is covered with
bacteriorhodopsin. When the plate is withdrawn at a certain speed, a monolayer of
rhodopsin adheres to the plate with all the molecules oriented in the same direction.
Repeating this process deposits a second layer, then a third, and so on. Information is
stored by assigning 0 to the green state and 1 to the orange state. Miniature lasers of the
type use thin fiber optic communications devices are used to switch between the two
states. Irradiating the disk with a green laser converts the green state to the orange state,
storing a 1, resetting the bit is accomplished by irradiating the same small area of the dusk
with a red laser. Data stored on the disk are read by using both lasers. The disk would be
scanned with the red laser and any bit with a value 1 would be reset using the green laser.
This is analogous to the way in which both magnetic and electrical memories are read
today, but with one important difference: "Because the two states take only five

Dept. of ECE 8
 PROTEIN MEMORY

picoseconds (five trillionths of a second) to flip back and forth, information storage and
retrieval are much faster than anything you could ever do magnetically or electrically,"
explained Birge. In theory, each pigment molecule could store one bit of information. In
practice, however approximately 100,000 molecules are sued. The laser beam as a
diameter if approximately 10 molecules and penetrates through the 1,000 molecule thick
layer. Although this reduces the amount of information that can be stored on each disk, it
does provide fidelity though redundancy. "We can have half the molecules or more in a
disk fall apart and there would still be enough excited by the laser at each spot to provide
accurate data storage," said Birge. And even using 100,000 molecules per data bit, an old
5.25 inch floppy disk could store well over 500 megabytes of data. Faster, higher-density
disk storage is a laudable goal, but the big stakes are in improving on semiconductor
components. Birge, for example, is developing a random access chip using the
bacteriorhodopsin system. Instead of having millions of transistors wired together on a
slab of silicon, there would be millions of tiny lasers pointed at a film of
bacteriorhodopsin. "These RAM chips would actually be a little bigger than what we
have," he said, "but they would still be 1,000 times faster because the molecular
components work so much faster than ones made of semiconductor materials."

Why Bacteriorhodopsin?

Bacteriorhodopsin (BR), a photon-driven proton pump, is a transmebrane protein

found in the purple membrane of the archaeon, Halobacterium halobium. The BR is
organized with polyprenoid lipid chains in the hydrophobic moiety on a highly ordered
two-dimensional lattice in the membrane. Enormous interest has been generated by
application of the structural and dynamic properties of the BR protein in bioelectronic
devices. Embedding bacteriorhodopsin in membrane films is essential to achieve
functionality of such devices. Although some prototype devices have been fabricated,
instability of the membranes is acknowledged as a major impediment to the development
of BR-based devices. Whereas there has been much biotechnological experimentation on
the properties and functionalities of the BR protein, comparatively little attention has been
paid to the critically important supporting lipid matrices in which BR functions. Archaeal
membranes consist predominantly of isoprenoid chains ether-linked to an alcohol such as
glycerol. These isoprene structures are unique and are not prone to decomposition at high
temperatures. That isoprene molecular fossils have been found in geologic Miocene

Dept. of ECE 9
 PROTEIN MEMORY

deposits attests to the extreme durability and stability of these structures. In viable
organisms BR functions in high efficiency at temperatures as high as 140˚C, in near-
saturating concentrations of salt, and in highly acidic environments. Specific lipids of the
purple membrane are required for normal bacteriorhodopsin structure, function, and photo
cycle kinetics. At ambient temperatures, the membranes are in a liquid crystalline state
that provides optimal functioning of the BR. Adaptations of the membrane lipids allow
maintenance of the liquid crystalline state even as the indigenous conditions change.
Nevertheless, as a consequence of light absorption, BR initiates a photo cycle through
structurally distinctive conformations, causing tractable optical and electronic properties
of the protein. The characteristics of the lipids in the membrane dictate the large-
amplitude motions of BR, and by consequence the broad utility of BR in bioelectronic
devices. The research to be developed under this topic would have military and civilian
importance. Some of the bionanotechnological applications that are anticipated include:
ultra rapid optical data acquisition with parallel processing capabilities and extreme high
density holographic three-dimensional data and image storage. Adapting natural
membranes or engineering alterations in those structures would have significant
advantages over the artificial membranes that are currently employed. The naturally
derived membranes would be biodegradable, eliminating the necessity for the disposal of
products that would be toxic or recalcitrant to decomposition. Another major advantage
would be the broad utility of the membranes with a variety of bioengineered
bacteriorhodopsin molecules. Using site directed mutagenesis it is known that the
molecular and physical properties of the protein molecule can be altered. Thus, with the
molecular biological tailoring of bacteriorhodopsin for specific phototropic properties, it
will be possible to optimize the photo cyclic intermediates for distinctive properties in
particular applications utilizing natural membranes.

Dept. of ECE 10
 PROTEIN MEMORY

Bacteriorhodopsin Optical Memory

1.Purple membrane from Halobacterium Halobium

2.Bitable red/green switch

3.In protein coat at 77K, 107-108 cycles

4.10,000 molecules/bit

5.Switching time, 500 femtoseconds

6.Monolayer fabricated by self-assembly

7.Speed currently limited by laser addressing

Using the purple membrane from the bacterium Halobacterium Halobium, Prof.
Robert Birge and his group at Syracuse University have made a working optical bistable
switch, fabricated in a monolayer by self-assembly, that reliably stores data with 10,000
molecules per bit. The molecule switches in 500 femtoseconds--that's 1/2000 of a
nanosecond, and the actual speed of the memory is currently limited by how fast you can
steer a laser beam to the correct spot on the memory.

Dept. of ECE 11
 PROTEIN MEMORY

The Structure of Bacteriorhodopsin

The retinal protein bacteriorhodopsin is the major photosynthetic protein of the

archaeon Halobacterium salinarum. It converts the energy of "green" light (500-650
nm, max 568 nm) into an electrochemical proton gradient, which in turn is used for
ATP production by ATP synthase. It functions as a light-driven proton pump,
transporting protons out of the cell, and exemplifies vectorial catalysis.

Bacteriorhodopsin - as all retinal proteins from Halobacterium - folds into a

seven-transmebrane helix topology with short interconnecting loops. The helices
(named A-G) are arranged in an arc-like structure and tightly surround a retinal
molecule that is covalently bound via a Schiff base to a conserved lysine (Lys-216) on
helix G. The cross-section of BR with residues important for proton transfer and the
probable path of the proton are shown in Fig.1. The 3D structure is also available.

Retinal separates a cytoplasmic from an extra cellular half channel that is lined
by amino acids crucial for efficient proton transport by BR (especially Asp-96 in the
cytoplasmic and Asp-85 in the extra cellular half channel). The Schiff base between
retinal and Lys-216 is located at the center of this channel. To allow vectorial proton
transport, de- and reprotonation of the Schiff base must occur from different sides of
the membrane. Thus, the accessibility of the Schiff base for Asp-96 and Asp-85 must be
switched during the catalytic cycle. The geometry of the retinal, the protonation state of
the Schiff base, and its precise electrostatic interaction with the surrounding charges
(Asp-85, Asp-212, Arg-82) and dipoles tune the absorption maximum to fit its
biological function.

Dept. of ECE 12
 PROTEIN MEMORY

Fig. 1: The 3D structure of bacteriorhodopsin (cross section of the structural

model). Selected residues important for proton transfer steps are marked. The probable
path of protons is indicated by arrows.

The photo cycle (catalytic cycle) of bacteriorhodopsin

Absorption of a photon by bacteriorhodopsin initiates a catalytic cycle that leads

to transport of a proton out of the cell. Several intermediates in the photocycle have
been identified by spectroscopic techniques (Fig.2). By application of a multitude of

Dept. of ECE 13
 PROTEIN MEMORY

biophysical techniques, the exact nature of the changes in each step of the cycle has
been determined and has been related to transport function.

Fig. 2: The photo cycle of bacteriorhodopsin

The cycle can be formally described in terms of six steps of isomerization (I),
ion transport (T), and accessibility change (switch S). Retinal first photo-isomerizes
from all-trans to a 13-cis configuration followed by a proton transfer from the Schiff
base to the proton acceptor Asp-85. To allow vectoriality, reprotonation of the Schiff
base from Asp-85 must be excluded. Thus, its accessibility is switched from extra
cellular to intracellular. The Schiff base is then reprotonated from Asp-96 in the
cytoplasmic channel. After reprotonation of Asp-96 from the cytoplasmic surface,
retinal reisomerizes thermally and the accessibility of the Schiff base switches back to

Dept. of ECE 14
 PROTEIN MEMORY

extra cellular to reestablish the initial state. These steps represent the minimal number
of steps needed to account for vectorial catalysis in wild-type bacteriorhodopsin.

The catalytic cycle step by step

Dynamic structural changes occuring in chromophore and protein during the

light-induced reaction cycle can be detected either directly by time-resolved
spectroscopic techniques (ultra fast laser spectroscopy, flash photolysis, ESR
spectroscopy, FTIR spectroscopy) or by trapping intermediate states, determining their
structures by static methods (NMR spectroscopy, electron microscopy, neutron
scattering) and comparing it with the ground state.

Primary reaction: the photoisomerization of retinal from all-trans to 13-cis

In a stereo selective process, all-trans retinal is photoisomerized to 13-cis

retinal. This process has been time-resolved to few femtoseconds. Within 500 fs, all-
trans retinal isomerizes to 13-cis retinal, resulting in J600 which is converted to K590
within another 5 ps.

From the K590 to the L550 intermediate

The K590 intermediate is transformed to the L550 intermediate within 2 µs. The
hydrogen bonding interaction in the extra cellular channel between the protonated
Schiff base and Asp-85, which involves a water molecule, is strengthened.

First proton translocation step: from L550 to M410(EC)

The M state is reached from the L state within several microseconds. This
transition involves transfer of a proton from the Schiff base to Asp-85 in the extra
cellular half-channel.

First accessibility switch reaction from extra cellular to cytoplasmic:

M410(EC) to M410(CP)

Dept. of ECE 15
 PROTEIN MEMORY

To allow vectorial proton transport, de- and reprotonation of the Schiff base
must occur from different sides of the membrane. This accessibility switch occurs at the
level of the M intermediate: M410 (EC) to M410 (CP). Thus, the originally described
"M" intermediate is in fact split into two or more different intermediates all having
yellow color.

Second proton transfer step: from M410(EC) to N560

Reprotonation of the Schiff base from Asp-96 in the cytoplasmic half-channel

occurs during transformation from the M410(EC) to the N560 intermediate within
milliseconds. Reprotonation of Asp-96 by a proton from the cytoplasm also occurs
during the lifetime of the N560 intermediate.
It should be noted that Asp-96 functions as proton storage for reprotonation of
the Schiff base. Therefore, the proton does not originate directly from the cytoplasm.
This detail solves the puzzling phenomenon that the transport rate of this proton
transporter is not pH-dependent (within limits).

Thermoisomeration of retinal from 13-cis to all-trans: N560 to O640

The transition of the N560 to the O640 intermediate is the thermal 13-cis to all-
trans isomerization of retinal in the environment of protonated Asp-96 and protonated
Asp-85.

Second accessibility switch reaction from cytoplasmic to extracellular:O640 to BR

Deprotonation of Asp-85 completes the catalytic cycle. Switching the

accessibility of the Schiff base back from extracellular to intracellular occurs within ca
5 ms and results in restoration of the initial state.

Dept. of ECE 16
 PROTEIN MEMORY

Site-directed mutagenesis of bacteriorhodopsin

An important tool for these studies, and also for spectroscopic investigations on
structure and dynamics, is the possibility to produce specifically modified proteins by
site directed mutagenesis and homologous over expression. By this method the role of
individual amino acid residues for the transport mechanism can be investigated, or
reporter molecules can be introduced at certain positions. Several mutants interfere with
the photo cycle and proton transport and may permit to trap intermediates of the cycle.

3-Dimensional Optical Memories

Three-dimensional optical memory storage offers significant promise for the

development of a new generation of ultra-high density RAMs. One of the keys to this
process lies in the ability of the protein to occupy and form cubic matrices in a polymer
gel, allowing for truly three-dimensional memory storage. The other major component in
the process lies in the use of photons to read and write data. As discussed earlier, storage
capacity in two-dimensional optical memories is limited to approximately 1/lambda2
(lambda = wavelength of light), which comes out to approximately 108 bits per square
centimeter. Three-dimensional memories, however, can store data at approximately
1/lambda3, which yields densities of 1011 to 1013 bits per cubic centimeter. The memory
storage scheme which we will focus on, proposed by Robert Birge in Computer (Nov.
1992), is designed to store up to 18 gigabytes within a data storage system with
dimensions of 1.6 cm * 1.6 cm * 2 cm. Bear in mind, this memory capacity is well below
the theoretical maximum limit of 512 gigabytes for the same volume (5-cm3).

Dept. of ECE 17
 PROTEIN MEMORY

Data Writing Technique

Bacteriorhodopsin, after being initially exposed to light (in our case a laser beam),
will change to between photo isomers during the main photochemical event when it
absorbs energy from a second laser beam. This process is known as sequential one-photon
architecture, or two-photon absorption. While early efforts to make use of this property
were carried out at cryogenic temperatures (liquid nitrogen temperatures), modern
research has made use of the different states of bacteriorhodopsin to carry out these
operations at room-temperature.

The process breaks down like this:

Upon initially being struck with light (a laser beam), the bacteriorhodopsin alters
its structure from the bR native state to a form we will call the O state. After a second
pulse of light, the O state then changes to a P form, which quickly reverts to a very stable
Q state, which is stable for long periods of time (even up to several years).

The data writing technique proposed by Dr. Birge involves the use of a three-
dimensional data storage system. In this case, a cube of bacteriorhodopsin in a polymer
gel is surrounded by two arrays of laser beams placed at 90 degree angles from each
other. One array of lasers, all set to green (called "paging" beams), activates the photo
cycle of the protein in any selected square plane, or page, within the cube. After a few
milliseconds, the number of intermediate O stages of bacteriorhodopsin reaches near
maximum. Now the other set, or array, of lasers - this time of red beams - is fired.

The second array is programmed to strike only the region of the activated square
where the data bits are to be written, switching molecules there to the P structure. The P
intermediate then quickly relaxes to the highly stable Q state. We then assign the initially-
excited state, the O state, to a binary value of 0, and the P and Q states are assigned a
binary value of 1. This process is now analogous to the binary switching system which is
used in existing semiconductor and magnetic memories. However, because the laser array
can activate molecules in various places throughout the selected page or plane, multiple
data locations (known as "addresses") can be written simultaneously - or in other words,
in parallel.

Dept. of ECE 18
 PROTEIN MEMORY

Data Reading Technique

The system for reading stored memory, either during processing or extraction of a
result relies on the selective absorption of red light by the O intermediate state of
bacteriorhodopsin. To read multiple bits of data in parallel, we start just as we do in the
writing process. First, the green paging beam is fired at the square of protein to be read.
After two milliseconds (enough time for the maximum amount of O intermediates to
appear), the entire red laser array is turned on at a very low intensity of red light. The
molecules that are in the binary state 1 (P or Q intermediate states) do not absorb the red
light, or change their states, as they have already been excited by the intense red light
during the data writing stage.

However, the molecules which started out in the binary state 0 (the O intermediate
state), do absorb the low-intensity red beams. A detector then images (reads) the light
passing through the cube of memory and records the location of the O and P or Q
structures; or in terms of binary code, the detector reads 0's and 1's. The process is
complete in approximately 10 milliseconds, a rate of 10 megabytes per second for each
page of memory.

Clearly, there are many advantages to protein-based memory, among the most
significant being cost, size, and memory density. However, there are still several barriers
standing in the way of mass-produced protein-based memories.

Dept. of ECE 19
 PROTEIN MEMORY

Data Erasing

To erase data; a brief pulse from a blue laser returns molecules in the Q state back
to the rest state. The blue light doesn't necessarily have to be a laser; you can bulk-erase
the cuvette by exposing it to an incandescent light with ultraviolet output.

Dept. of ECE 20
 PROTEIN MEMORY

Refreshing the memory

In Protein memory the read/write operations use two additional parity bits to guard against
errors. A page of data can be read nondestructively about 5000 times.Each page is
monitored by a counter and after 1024 reads, the page is refreshed by a new write operation.

Merits

• This is based on a protein that is inexpensive to produce in quantity

• .The system has the ability to operate over a wider range of temperatures than the
semiconductor memory.

• The data is stable, ie the memory system’s power is turned off, the molecules retain
their information.

• This is portable, ie we can remove small data cubes and ship GBs of data around for
storage or backups.

Dept. of ECE 21
 PROTEIN MEMORY

Conclusion

Birge’s system, which he categorizes as a level-1 prototype (i.e., a proof of concept),

sits on a lab bench. He has received additional funding from the USAir force, Syracuse
University, and the W.M.Keck foundation to develop a level-2 prototype. Such a prototype
would fit and operate within a desktop personal computer. “We are a year or two away from
doing internal testing on a level-2 prototype”, says Birge. “Within three or five years, we
could have a level-3 beta-test prototype ready, which would be a commercial product.”

Dept. of ECE 22