Security Printing Speech

Back to Speech Index


DNA Based Steganography for Security Marking


XIX INTERNATIONAL SECURITY PRINTERS' CONFERENCE
Montreux, 14-16 May 2003

Wendell M. Smith, Technology Transfer Group

ABSTRACT

DNA the magic code of life, the form of which was identified in 1953 as a double helix of deoxyribonucleic acid (DNA), is made up of two strands of four bases in varying triplets that repeat over and over in a very long molecule. It is a very complex code and can be used to identify specific human beings and to clone animals. DNA is also being used as an information carrier and as an encryption and computation device. DNA has recently been used to make an "ink" or taggant for security printing.

Now the invention of combining the DNA ink idea with Steganography, the art of hiding authentication information inside biological information, makes a good idea into a foolproof idea. With the DNA LockTM technology, there is no way to discover the hidden DNA code in the mixture without knowing the primer key code needed for the PCR analysis to identify the authentic ink. The most non-counterfeitable mark ever conceived has been made possible by this development.

The inventors have made several sample inks, printed microdots with a standard printer, sent the messages out in the mail, retrieved them and been able to confirm the validity of the ink employed.


Summary

DNA the magic code of life, the form of which was identified in 1953 as a double helix of deoxyribonucleic acid (DNA), is made up of two strands of four bases in varying triplets that repeat over and over in a very long molecule. It is a very complex code, the human DNA is over 3 billion characters long. It can be used to identify specific human beings and has been used to clone animals. In a less dramatic but important application, DNA is being used as an information carrier and as an encryption and computation device. The DNA code could theoretically carry a whole written novel in one single molecule. DNA has recently been used to make an "ink" or taggant for security printing.

Now the invention of combining the DNA ink idea with Steganography, the art of hiding authentication information inside biological information, makes a good idea into a foolproof idea. With the DNA LockTM technology, there is no way to discover the hidden DNA code in the mixture without knowing the primer key code needed for the PCR analysis to identify the authentic ink. The most non-counterfeitable mark ever conceived has been made possible by this development. The inventors have made several sample inks, printed microdots with a standard printer, sent the messages out in the mail, retrieved them and been able to confirm the validity of the ink employed.


Background

The inventors, Dr. Carter Bancroft and Dr. Catherine Clelland of the Mount Sinai School of Medicine have been working in this field for many years. They were working on the computational capabilities of the DNA code. They saw the opportunity to improve on security marks in use today by putting Steganographic concepts to work in the DNA field. On November 6, 2001, US patent # 6,312,911 DNA Based Steganography was granted and assigned to the Mount Sinai School of Medicine. The name DNA LockTM is the trademark adopted for use with this technology.

At the same time, Technology Transfer Group was actively making licensing agreements in the security printing field, particularly in holography, and developing many contacts with this very special industry. TTG now represents several companies' technologies and has developed business connections of various kinds throughout the world. We have presented papers before this group on the subjects of High Speed on Press Hologram embossing and application (NovaVision), Digital foil printers - applying holograms digitally with numbering or coding (Impress Systems), and Fluorescent Inks and systems (PhotoSecure). When we learned of DNA based Steganography, we moved quickly to obtain the worldwide rights to the patents.


Present use of DNA in security

Several companies are printing security marks on to labels and documents using DNA inks. Ones that come to mind include DNA Technologies, November AG, Smartwater, Trace Tag Int'l, and Hardy Wines. Some paper companies have put DNA taggants in the paper.

These DNA marks or taggants are usually added to inks along with some other security device since most of the DNA inks are only verifiable in a laboratory situation. It is usually desirable to have an overt easily noticed mark for on site verification. Covert marks are usually observed by fluorescent taggants or holograms in conjunction with the DNA mark. In the case of November AG, the company uses a DNA taggant only and has developed a marker pen that contains a mating single strand of DNA. When (if) the mating strand sees the original DNA, it combines and the marker pen ink fluoresces. Since the DNA marks employed by these companies are not "hidden", the DNA itself can be recovered from a marked object and could be copied. The DNA Steganography patents will make any of these processes much better by hiding the secret DNA code inside a complex "soup" of DNA codes. Only the intended user can discover and read the authentication DNA code.

What is DNA?

All forms of life, including humans, contain a molecule that carries the genetic code of that life, DNA. The form of the molecule deoxyribonucleic acid, known as DNA, was identified in 1953 (yes we are happy to be celebrating the 50th anniversary of this momentous event this year!!) by Watson and Crick is a double helix made up of two strands of four bases Adenine, Guanine, Cytosine and Thymine. These bases are found in triplets of varying sequence repeated over and over in a very long and complex code. Each polymer strand is held together by hydrogen bonds. In the linking of the two strands, A will bond to T only and C to G only. The DNA molecule in a single cell is twisted together so densely that it can only be seen by a very powerful microscope; yet it would reach over 6 feet if it could be stretched out fully.

Scientists have been able to identify differences in human DNA that show up in people with various maladies like cancer, heart disease etc. and are even finding differences that relate to personality and mental activity. Knowledge of this complicated and unique code is now leading to wonders of medical science as drugs are being developed to work around the DNA code. This field is as controversial as the development of atomic energy with equal scientific and social consequences. It has proven possible to place a selected DNA code into an animal and give the animal almost any properties you choose.

None of that makes any difference to us in the security printing field since we do not and would not use live animals or human DNA for any reason whatsoever. The DNA for inks and taggants can be a combination of cheaply manufactured synthetic DNA, plus bulk DNA from processed foods (Hardy wines uses their own wine for the DNA source), or from other plants or animals.

What is Steganography?

Steganography is the art of hiding specific information inside a great excess of similar-appearing information. The microdot is a means of concealing messages steganographically was developed by Professor Zapp (note 3) and used by German spies in the Second World War to transmit secret information. A microdot ("the enemy's masterpiece of espionage") was a greatly reduced photograph of a typewritten page that was pasted over a full stop (a period) in an innocuous letter.

In another field we see an example of steganography in the current children's book "Where is Waldo". The cover photo helps demonstrate steganography, finding Waldo is a challenge because he is there but how do you identify which of the many characters that all look alike is really Waldo?

Classical steganography involves hiding a message within numerous similar "dummy" messages. It has been used to transmit secret messages since ancient times, even predating Caesar's cipher. Empirical modern proof of the soundness of the technique is the recent use of images and audio watermarking in the digital world for anti-piracy, including copyright protection by Microsoft, AT&T, and IBM, in systems such as IBM™ Cryptolope, SysCoP, FASTAudio, TALISMAN, ACCOPI, and KryPict (Europe). Steganography is a preferred protection method because it requires virtually no pre-or post-processing, and affords both real-time processing and variable levels of accessibility to a hidden full text. The challenges for the technique are resistance to image and sound processing techniques, such as filtering, compression, resampling, etc.

Conventional steganographic cryptanalysis attack can operate in basically two modes. The first is to attack the signal itself to remove the watermark(s). This does not require knowledge of the mark or its location, and proceeds by rotating, scanning, compressing, or otherwise modifying the ciphertext containing the plaintext (Note 1). The second kind of attack is cryptologic. Examples of cryptologic include: 1. Social engineering (espionage to observe the sender hiding the information, and thereby determine the shape and position of the watermark); 2. Traditional attacks on portions of the ciphertext such as frequency/entropy measurements; and 3. Attacks on the integrity of the ciphertext (chopping or tessellating) to remove automatic detection of watermarks. However, as described below, none of these modes of attack makes sense in the context of DNA-based steganography, nor do they represent a threat to the technique. Instead, a cryptanalysis of DNA-based steganography/watermarking requires a biochemical approach consistent with the biochemical, DNA-based nature of the highly concealed marks.

Steganography applied to DNA ink systems for security printing creates a truly non- counterfeitable mark. And it is not "hackable" in the computer sense in that it is a biological marker, not a mathematical or computer generated number. No large number of supercomputers working together can discover which of the multitude of DNA AGCT polymer chains is the secret one since computers can't do biochemical experiments.


What is DNA based Steganography

The technology described in this proposal is based upon a "genomic Steganography" procedure that was developed and published by Clelland et al, 1999 (note 2). That paper, describes a doubly steganographic concealment technique for hiding a secret message encoded in DNA. A DNA molecule is constructed containing an encoded message flanked by PCR primers keys. The DNA-encoded message is first camouflaged within the enormous complexity of the genomic DNA of an organism (human DNA, a combination of genomes from different species, or alternatively a synthetic, highly complex random DNA mixture). This results in the molecule being hidden by millions of other similar-looking DNA molecules, analogous to a DNA needle in an enormous DNA haystack.

The message is then further concealed by confining this DNA sample to a microdot the size of a period. The encoded message can then be recovered only by an intended recipient, who is able to find it, and more importantly, who knows the sequences of the PCR primers, or "keys" to the readout procedure. For readout, the recipient employs the PCR keys in a polymerase chain reaction (PCR), a highly sensitive, core technique in molecular biology that results in the production of exponential copies of the encoded message DNA molecule, allowing the molecule to be detected and then read by DNA sequence analysis.

A prototypical 'secret message' DNA strand contains an encoded message flanked by polymerase chain reaction (PCR) primer key sequences. Encryption is not of primary importance in steganography, so we can use a simple substitution cipher to encode characters in DNA triplets. Because the human genome contains about 3210 X 10X9 nucleotide pairs, fragmented and denatured human DNA provides a very complex background for concealing secret-message DNA. For example, a secret message 100 nucleotides long added to treated human DNA at one copy per haploid genome would be hidden in a roughly three-million-fold excess of physically similar DNA strands. Confining such a sample to a microdot might then allow even the Medium containing the Message to be concealed from an adversary.

However, the intended recipient, knowing both the secret-message DNA PCR primer sequences and the encryption key, could readily amplify the DNA and then read and decode the message.

Even if an adversary somehow detected such a microdot, it would still prove extremely difficult to read the message without knowing the specific primer key sequences. If 20-base random primers were used to amplify the DNA, separate amplifications with more than 1020 different primer pairs would be required, even allowing three mismatches per primer, followed by analysis of any PCR products obtained. Similar considerations apply to attempts to shotgun-clone the DNA sample and analyze the resultant clones. Even if the same primer pair was used on several occasions, an enemy trying to detect the primer sequences would face an extremely difficult experimental barrier. Further mathematical and biochemical analysis would therefore be expected to prove that the primer pairs used in this technique are not analogous to a classic, single-use, cryptographical "one-time pad".

Attempts by an adversary to use a subtraction technique to detect the secret-message could be blocked by using a random mixture of genomic DNAs from different organisms as background. The intended recipient could still use the same procedures to amplify and read the secret-message DNA, even if ignorant of the random mixture composition, and even if the primers artefactually amplified a limited number of genomic sequences, the encryption key would reveal which PCR product encodes a sensible message. This technique would also allow single or duplicate microdots to be used to send individual secret messages to each of several intended recipients, each of whom would use a unique set of primers to amplify only his or her intended message.

How do you make an ink or taggant?

DNA, when isolated and reduced in a laboratory, is a white (effectively colorless) solid. It is water soluble so it can be added to various vehicles for liquid form printing. It has so far been added to "standard" inks (security inks) and printed with a standard ink jet printer. Tests have also shown that it can be added to solvent based inks and it is anticipated that it can be added to nearly any ink at all without changing the printing characteristics in any noticeable way.

Very small amounts of DNA are adequate for the analysis if the location is known and if removal of the sample is possible.

Lifetime of the ink is expected to be unchanged by the DNA taggant. DNA is very long lasting and the modifiers and degraders are well know and uncommon in normal circumstances. DNA is stable for thousands, even millions, of years (note 6) and is being used to learn more about ancient people and animals. DNA is an extremely stable molecule and is thus ideal for use as a security label.

Experiments have shown that DNA is both resistant to degradation over millions of years (Note 6). In addition, the ability to perform downstream reactions on DNA molecules, such as PCR, is not affected by subjecting the DNA to extreme conditions of heat and pressure, and even harsh solutions such as bleach (note 7), none of which block the use of DNA for marks. It is anticipated that we can develop a DNA Mark that will be placed under a common seal, such as the security seal employed over the top of a bottle of pills. The placing of our DNA Lock mark there should not provide overt clues to the presence of the DNA, and should ultimately allow for a stable and efficient readout procedure.

To investigate the feasibility of this ink-marking scheme, we synthesized a secret-message DNA oligodeoxynucleotide containing an encoded message 69 nucleotides long flanked by forward and reverse PCR primers, each 20 nucleotides long. We prepared a concealing DNA that is physically similar to the secret-message DNA by sonicating human DNA to roughly 50 to 150 nucleotide pairs (average size) and denaturing it. We pipetted 6 ml of each solution containing 225 ng of treated human DNA, plus various amounts of added secret-message DNA, over a 16-point full stop (a period). It was printed on filter paper where it finally occupied an area about 20 times the size of the full stop.

Excision of the printed full stops, each containing about 10 ng of DNA and with a cross-section that was about 75% larger than a full stop on this page, yielded DNA microdots. Primer keys designed to amplify the secret-message DNA were used to perform PCR directly on DNA microdots, without prior DNA solubilization.

The products were analyzed by gel electrophoresis. An unamplified sample containing secret-message DNA yielded only a faint continuous smear. In contrast, amplification of DNA microdots containing either 100, 10 or 1 copies of the secret-message DNA per haploid genome, lanes 3–5, each yielded a single product of the expected size (arrow). No such product was detected in lane 6 or, lane 7 indicating a detection limit of about one secret-message DNA strand per haploid human genome.

The amplified band in lane 4 of (arrow) was excised, subcloned and sequenced. Use of the encryption key to decode the resultant DNA sequence yielded the encoded text, containing probably the most significant secret of the original microdot era: "June 6 invasion: Normandy".

In preliminary experiments, microdots containing 100 copies of secret-message DNA per human haploid genome, which had been attached using common adhesives to full stops in a printed letter, and posted through the US mail, yielded the correct PCR amplification product. Our technique could therefore be used in a similar way to the original microdots: to enclose a secret message in an innocuous letter. It should be possible to scale up the encoded message from the size of our simple example, perhaps by encoding a longer message in several smaller DNA strands. It should also be possible to use smaller microdots, which could be used for a variety of purposes, including cryptography and specific tagging of items of interest.

How do you confirm the mark?

DNA testing is a laboratory process. The ultimate testing is "sequencing"; actually identifying every one of the DNA base subunits in their naturally occurring pattern. For the DNA code that is highly concealed in DNA Steganography, the confirmation begins with a process called PCR (polymerase chain reaction) that can amplify and help identify a certain sequence by amplifying the very small amount of that DNA amongst the enormous "noise" of other DNA sequences. In medical studies, PCR is used to isolate specific human genes. In DNA Steganography, PCR is used to recover the secret DNA code that is hidden in an enormous background of concealing DNA. But only an authorized person who knows the proper PCR primer key sequences can recover the secret DNA code.

The PCR test uses both temperature cycling, as the two strands separate at a precise temperature, and new DNA synthesis. By doping the sample with the correct PCR primer keys, and using many rounds of PCR, the greatly hidden secret DNA code goes from a trace contaminant to the major DNA sequence present. The PCR step by itself yields a level of authentication. But, for absolute verification, the sequence of the secret DNA code can be determined, yielding the authentication code. But recall that only an authorized person can carry out this procedure.

For each of these procedures, fluorescent signals are employed. The PCR test equipment costs can cost as little as $25,000, and the test takes about two hours. The sequencing equipment costs roughly $40,000, and this test also takes only a few hours.

Many companies are working on faster and simpler PCR-based test equipment, especially since PCR testing is the presently preferred way of looking for anthrax and other bioterrorism agents. These companies will no doubt greatly reduce the time and cost of PCR-based verification of security marks, and the size of the equipment required. We do not however assume that DNA will be used solely on a security device any time soon since it is not an overt mark and cannot provide a quick response verification step.

What are the practical applications of DNA Inks?

We differentiate the security markets into four categories; negotiable instrument anti-forgery, personnel identity and access control, anti-theft marking, and product authentication. There is a blurring of these factors as time goes on and some companies are preparing combined answers to a whole business, a factory or whole product line. Some suppliers are finding ways to use their various technologies in all or several sectors of the market. There is an increasing need for a comprehensive organization-wide or global product-wide planning system.

DNA Steganography can be used in any version of a mark or taggant. As the ultimate, most foolproof mark, it can be used for each and every purpose – with the caveat that at present it will not be appropriate for an immediate verification step alone and must be part of a combination of elements in a device or document. There are several reasons why security devices are combining more than one feature, and we see that the best device will have DNA Steganography technology for the "ultimate" forensic verification step.

Early tests have been aimed at the product authentication market and we have made inks and printed documents, retrieved the marks and determined the original code. These printed marks are digital in nature permitting unique machine-readable marks for the all-important track and trace function. Many security marks are finding a secondary (sometimes primary) reason to exist as the automated input of information about the product or person that starts the data entry for the company's computer driven production and distribution activity. The security mark may be the very first entry into the IT software system and drive the whole system.

What are the next steps in development?

Three projects that will have significant impact on the success of DNA Steganography are: compatibility with various substrates, shelf life of inks and marks, and the hardware for reading the marks quickly in the field.

Substrates: there are many different papers and materials that security marks are commonly applied to/printed on. The inks/taggants need to be tested for compatibility with the many substrates appropriate to the various applications. The drug bottle, the dress label, the banknote, the credit card, the lottery ticket and so on. Some of these materials may need an overlay or adhesive mark attachment. Many may be printed in a normal manner. It is possible that a DNA mark that will be placed under a common seal, such as the security seal employed over the top of a bottle of pills. The placing of our DNA Lock mark there should not provide overt clues to the presence of the DNA, and should ultimately allow for a stable and efficient readout procedure.

Shelf Life; the DNA materials are stable and long lasting. Mixing it into an ink is not expected to change that fact, but clear evidence by testing is proposed. The inks need to be tested for shelf life before use and after printing.

Hardware; Rapid, inexpensive field-testing equipment is need and it needs to be simplified and incorporated into an ink/substrate/readout system. There are many companies working on this since DNA testing is and will continue to be a very important forensic activity for many medical and security reasons. When we will have a low cost fast reader in the field? Not real soon, but not too far distant either.


WMS March 14, 2003


REFERENCES:

Patents
  • DNA Based Steganography. F.C. Bancroft and C. Clelland . U.S. Patent # 6,412,911 issued 11/6/01
  • DNA-Based Computer. F. Guarnieri and F.C. Bancroft. U.S. Patent #5,955,322, issued 9/21/99
Books and Articles
  • Time Magazine, February 17, 2003:
    A Twist of Fate - by Michael Lemonick
    The Secret of Life - by Nancy Gibbs

  • The Double Helix: James Watson 1968 Simon and Schuster
  • Guarnieri, F., M. Fliss, and C. Bancroft. Making DNA add. Science 273: 220-3 (1996)
  • Guarnieri, F., M. Orlian, and C. Bancroft. Parallel operations in DNA-Based computation. Proceedings of DNA Based Computers III (1997), H. Rubin and D. H. Wood, Eds. (American Mathematical Society, DIMACS Series, Vol. 48), pp. 85-100. 399:533-4 (1999)
  • Bloom, B., and C. Bancroft. Liposome mediated biomolecular computation. Proceedings of DNA Based Computers V (1999), E. Winfree and D. K. Gifford, Eds. (American Mathematical Society DIMACS Series, Vol. 54), pp. 63-72
Notes
  1. ciphertext : reviewed in Katzenbeisser and Petitcolas, 1999:
  2. Clelland, C. T., V. Risca, and C. Bancroft. Hiding messages in DNA microdots. Nature. (Supplementary information is available on Nature's World-Wide Web site (http://www.nature.com) or as paper copy from the London editorial office of Nature)
  3. Kahn, D. The Codebreakers (Scribner, New York, 1996)
  4. Hoover, J. E. Reader's Digest 48, 1–6 (April 1946)
  5. Clayton, P. T. et al. Arch. Dis. Child 79, 109–115 (1998)
  6. Bancroft, C., T. Bowler, B. Bloom, and C. Bancroft. Long-term storage of information in DNA. Science 293:173-1765 (2001)
  7. Hall et al., 2003
COMPANY REFERENCES:
  • Microsoft (www.research.microsoft.com/security)
  • PhotoSecure (www.photosecure.com)
  • Impress Systems (www.impresssystems.com)
  • DNA Technologies (www.dnatechnologies.com)
  • Smartwater (www.smartwater.co.uk)
  • November AG (www.november.de)
  • Trace Tag Intl (www.tracetag.com)
  • Hardy wines (www.hardywines.com)
  • Polestar Ltd. (www.polestarltd.com)
  • Technology Transfer Group (earlier speeches: India, Seville, Spain, Sorento, Italy) (www.polestarltd.com/index.html)
CoAuthors
  • Catherine Taylor Clelland
  • Carter Bancroft
  • Department of Physiology and Biophysics,
  • Mount Sinai School of Medicine, New York,New York 10029, USA e-mail: cbancro@smtplink.mssm.edu

Author:
Wendell M. Smith
Technology Transfer Group
polestar@logic.bm
 


Back to Speech Index



© Copyright Polestar Ltd. 2008