Hiding data in DNA of living organisms

doi:10.4236/ns.2009.13023

Paper Menu >>

Journal Menu >>

Vol.1, No.3, 181-184 (2009)

doi:10.4236/ns.2009.13023

SciRes

Natural Science

Hiding dat a in DNA of living organisms

Shu-Hong Jiao1, Robert Go utte 2

1Information and Telecom Department, Harbin Engineering University, Harbin, China

2Universite De Lyon , Lab Creatis Lrmn, INSA Lyon, France; [email protected]

Received 12 June 2009; revised 16 July 2009; accepted 18 July 2009.

ABSTRACT

Recent research has considered DNA an inter-

esting medium for long-term and ultra compact

information storage and a stegomedium for hid-

den messages. Artificial components of DNA

with encoded information can be added to the

genome of living organisms, such as common

bacteria. With this approach, a medium for very

height densities information storage, water-

marks for protection patents of genetically mo-

dified organisms (GMOs) and secure public

keys for d ecryp ting hidd en infor mati on in stega-

nocryptography, can be obtained. In this paper,

we have selected a Bacillus subtilis gene (tatAD)

and use the specific properties of silent muta-

tions to obtain a biologically innocuous product.

An adapted code for the message insertion in

this gene is proposed.

Keywords: DNA; Genome; Coding; Steganography;

Living Organism

1. INTRODUCTION

There has been growing interest in using DNA to store

information, one the main reasons being the very high

storage densities that can be achieved. The durability

of DNA would make it particularly useful for preserving

archival material over extensive periods of time (long-

term storage). Message DNA has been used in computa-

tions in biologic mathematics, in

steganography and as a

mean of short-term trademarks. These different fields

use the specific properties of genetic code and the pos-

sibilities to encode artificial information.

However, the artificial introduction of a new gene or the

modification of an existing gene in the DNA of a living

organism, can involve prejudicial deterioration of the beha-

vior and the reproduction of this species. The solution sug-

gested must thus notably allow a great computer security,

but also the genetic consequences of their implementation.

2. THE GENETIC CODE

The genetic code is the biochemical basis of heredity and

nearly universal in all organisms (eukaryotes or pro-

karyotes): humans, animals, plants, bacteria and vi-

ruses.

DNA (deoxyribonucleic acid) is a long molecule,

with two strands rolled up in a double helix. Each strand

is formed by sugar phosphate backbone, connected with

single molecules called bases, which contain carbon, ni-

trogen and cyclical structures. The four bases are known

as adenine A, thymine T, guanine G and cytosine C.Any

strand of DNA adheres to its complementary strand, in

which T substitute for A, G for C, and vice versa. The

links between pairs of bases are responsible for binding

together two stands together to form the double helix.

The genetic information in DNA is found in the

ordered sequence of these four bases (the double helix

structure of DNA introduces redundant information,

resulting in complementary nitrogenous base links).

Unlike DNA, RNA (ribonucleic acid) is a single

stranded molecule and does not form of double helix.

The bases are the same that DNA, except U (uracil),

which replaces T. The complementarily becomes:

U ↔ A and G ↔ C

Messenger RNA (mRNA) carries all the genetic infor-

mation to ribosome for proteins synthesis by the cell.

A codon is a triplet of three bases (T,C,A and G). With

these four letters, 43═64 combinations are possible. With

three exceptions, each codon encode for one of the 20

amino acids, used in the proteins synthesis (the three ex-

ceptions are TAA, TAG and TGA: codons STOP).

ATG correspond at methionine within the gene; at the

beginning of the gene, ATG is also a signal START. Ribo-

some assembles individual amino acids into peptide chains.

Peptides are short chains of amino acids that are linked

together. If the number of amino acids in the chain reaches

around ten or more, such substances are called polypep-

tides and large polypeptides are called proteins.

S. H. Jiao et al. / Natural Science 1 (2009) 181-184

182

Figure 1.1. The genetic code (DNA).

1) In yellow: The eight blocks with a single amino acid.

2) Blocks repaired with only two first bases.

Figure 1.2. Simplified code.

3. DEGENERACY OF GENETIC CODE

Many codons are degenerate, or redundant, meaning two

or more codons (synonymous codons) may code for the

same amino acid. See Figure 1.1. Degenerate codons

typically differ in their third position. (e.g.: GAA, GAC for

glutamine). See Fi gur e 1.2.

The degeneracy of the genetic code is what accounts for

the existence of “silent mutations”. A mutation occurs

when a DNA gene is damaged or changed. A silent muta-

tion is a mutation that does not alter the amino acid of a

gene, usually because of codon ambiguity.

4. STATE OF ART, PROPOSED CODE

AND SPECIFIC PROPERTIES

The first paper on hiding messages in DNA was published

by T. Clelland and al (ref 1) in 1999 and involved the in-

sertion of a brief message in a sample of human DNA. In

2001, these same authors published a paper showing of

possibilities of long-term storage of information in DNA

and used a classical codon encoding for the English al-

phabet. In 2003, C. Smith and al (ref 3) described a possi-

ble code for encrypting data.

In DNA and Boris Shimannovsky and al (ref 4) pro-

posed an interesting and original arithmetic code. Pak

Chung Wong and al (ref 5) presented new potential appli-

cations for DNA organic data memory.

The advantages of these approaches in production keys

were presented by Masanori Arita (ref 6) and Tanaka and

all, in 2005 (ref 7). Recently, Nozomu Yachie and al (ref 8)

presented a very complete methodology, simple, flexible

and robust of data storage based on sequence alignment of

genomic DNA of living organism. D. Heider and al publish,

in 2007, a program called DNA Crypt whose use is centred

on the patent protection of genetically modified organisms

(GMOs) (ref 9).

In this work we propose a method of coding that satis-

fies two conditions:

Limitation of changes in the gene marker

Possibility to transfer the encrypted message in a key,

made with a polypeptide chain.

5. CHOICE OF SITE

Bacillus subtilis is a non-pathogenic bacterium, which

form pores that can survive in extreme conditions. This

bacterium is a model which has already been chosen as

storage of information by several teams.

To illustrate our proposed code, we have selected a

component of this genome: the gene tatAD (alternate name

ycbz). See Figure 1.3. This gene is very short; its length is

210 base pairs, which correspond to 70 amino acids. It is a

protein classified stable.

Figure 1.3. Bacillus subtilis.

S. H. Jiao et al. / Natural Science 1 (2009) 181-184

183

The following table lists the genetic code of these 70

amino acids

ATG TTT TCA AAC ATT GGA ATA CCG GGC

TTG ATT CTC ATC TTC GTC

ATC GCC ATT ATT ATT TTT GGC CCT TCC AAG

CTG CCG GAA ATC GGG

CGT GCC GCG AAA CGG ACA CTG CTG GAA TTT

AAC AGC GCC ACA AAC

TCA CTT GTG TCT GGT GAT GAA AAA GAA GAG

AAA TCA GCT GAG CTG

ACA GCG GTA AAG CAG GAC AAC AAC GCG GGC

A short list concern only the underlining codons: TCA

GGA CCG GGC CTC ……, codons-owned blocks of

the Table 2, in yellow mark.

6. IMPLEMENTATION

6.1. Encryption

By example, we consider the message “CODING”. After

translation with ASCII code (with 8 bits per character), this

message requires 48 bits.

● C 0 1 0 0 0 0 0 1

● O 0 1 0 0 1 1 1 1

● D 0 1 0 0 0 1 0 0

● I 0 1 0 0 1 0 0 1

● N 0 1 0 0 1 1 1 0

● G 0 1 0 0 1 1 1 1

With the rate of 2 bits per codon, this message requires

24 codons. For his implementation, only 35 codons are

selected (codons underlined).The beginning and the end of

this message necessarily belongs to these 35 codons, so

that only two 6 bits numbers (two times three codons) are

needed for their localization.

If we assume that the beginning is located at the 9th

codon, and is 48 bits long (24 codons), the end of the mes-

sage will be located at 32 th codon.

Beginning at 9→ 00 10 01

End at 32→ 01 00 00

Six codons are needed for their localization.

/1 2 3/-//4 5 6//7 8///9 10 11 12 13 14 15 16 17 18 19 20 21

22 23 24 25 26 27 28 29 30 31 32///33 34 35.

The coding consists in replacing the 3rd base of these

codons by A, C, G, T according this table:

A if 00 C if 01

G if 10 T if 11

Before coding we, in the first line have

ATG TTT TCA AAC ATT GGA ATA CCG GGC TTG

ATT CTC ATC TTC GTC

After coding we obtain:

ATG TTT TCA AAC ATT GGG ATA CCC GGG TTG

ATT CTA ATC TTC GTA

The first 24 codons selected on the short list are:

CCT TCC CTC CCG GGG CGT

GCC GCC CGG ACA CTG CTG

GCC ACA TCA CTT GTG TCT

GGT TCA GCT CTG ACA GCG

After coding, we obtain:

CCC TCA CTA CCT GGC CGA

GCT GCT CGC ACA CTC CTA

GCC ACA TCG CTC GTC TCA

GGT TCG GCC CTA ACC GCT

4 codons (underlined) are not modified by the coding.

Only 25 codons of the original gene’s 70 codons (tatAD)

are modified. All the mutations are particular missense

mutations: silent mutations; with its location, the encryp-

tion of this message requires 60 bits and 25 mutations.

Thus, on average, each silent mutation can insert slightly

more than 2 bits of hidden message.

6.2. Decryption

Decryption is particularly easy. After obtaining the begin-

ning and the end of the message from the first 6 codons of

the shortlist, it is sufficient to replace, in the sequence of

codons corresponding to the message, the last base of each

codon by its equivalent in bits:

(A→00, C→01, G→10, T→11)

In this example, a very short gene is used for demonstra-

tion purposes, but this method can be applied to longer

genes, that are more than 10000 base pairs long (such as

the gene srfAA de Bacillus subtilis) which then allows the

insertion of messages 50 that are time longer (~2400 bits

and therefore 300 ASCII characters).

7. MATERIALISAT ION OF THE KEY,

COMPLEMENTS

7.1. Producing the Key

It is possible to replace the list of 30 codon, which can be

synthesized and materialized in the form of a polypeptide,

as in the preceding example, with a new list of 60 amino

acids. Indeed, if we consider the beginning of the Short list

(after coding):

TCA GGG CCC ……..We have the possibility to intro-

duce a peptide chain:

Ser (TCA) Gly (GGG) Pro (CCC) ….

But it is not possible, with only this chain, to redis-

cover the different codons of the short list in reason of

the redundancy of the genetic code.

Ser correspond to TCT TCC TCA TCG

Gly to GGT GGG GGA CGG and

Pro to CCT CCC CCA CCG

For resolve this difficulty we propose the following

connections:

T C A G G G C C C

/\ /\ /\

T C X C A X’ G G X G G X’ C C X CCX’

If X =A and X’=C we obtain:

S. H. Jiao et al. / Natural Science 1 (2009) 181-184

184

8. CONCLUSIONS

T C A C A C G G A G G C C C A C C C

Ser His Gly Gly Pro Pro

new peptidic chain. With the chain we can retrieve the

three codons of the short list

The use of the degeneracy of the genetic code and, in par-

ticular, the silent mutations, produces coding that does not

practically alter the properties of the inserted gene, nor the

characteristics of the host genome (very important condi-

tions when we working with the live organisms). Memori-

zation of the key information and the production of the

hidden message in the form of a physical polypeptide pro-

vide additional data transfer security, while the coding

protocol is being implemented).

For this, it is sufficient to conserve uniquely two

bases/in each codon.

T C C A G G G G C C C C and to regroup:

T C G G C C

C A G G C C

  

T C A G G G C C C

DNA is a storage medium extremely effective: it is com-

pact and his signature is innocuousness and secrecy. In the

spore form, Bacillus subtilis is resistant to extremes terres-

trial and extraterrestrial environment during the interstellar

travel, for example (ref 10). In the nearly future, this bio-

engineering method can be used in routine, for protection

of the gene patents, digital copyrights and tool for the

marking in biodiversity

for obtain the three original codons.

7.2. Comp lem ents

It is possible to associate this encryption method with

a binary encryption algorithm (AES, RSA, Blow-

fish). In this case, the inserted message is not the clear

version, but a secure version after using these specific al-

gorithms.

7.3. Simplified Version REFERENCES

If, in spite of the precaution to using silent mutations dur-

ing the identification of genetically modified organisms,

residual biological changes still exist in the host organism,

then a simplified version of the method discussed above

can be used.

[1] C. C. Taylor, V. Risca, and C. Bancroft, (1999) Hiding

messages in DNA microdots. Nature, 399, 533-534.

[2] C. Bancroft, T. Bowler, B. Bloom, and C. C. Taylor

(2001) Long-term storage of information in DNA. Sci-

ence, 293(5536), 1763-1765.

From the gene initially modified by the agrochemical

firm (if this gene is accessible to the user), we propose to

use this gene as stegomedium, to produce a custom key.

This key can allow access to the reference of manufacturer.

With this goal in mind, in the preceding example (with the

message “CODING” of 48bits), we select 24 codons of the

70 existing in the tatAD gene.

[3] G

. C. Smith, C. C. Fiddes, J. P. Hawkins, and J. P. L. Cox,

(2003) Some possible codes for encrypting data in DNA.

Biotechnology, 25 , 1125-1130.

[4] B. Shimanovsky, J. Feng, and M. Potkonjak, (2003)

Hiding data in DNA, LNCS 2578, 373-386.

[5] P. C. Wong, K. K. Wong, and H. Foote, (2003) Organic

data memory using the DNA approach, Communication

of the ACM, 46(1).

Each codon selected is identified by its rank in the chain

(7 bits per codons are necessary, here corresponding to a

total of 24x7=168 bits). These 168 bits are the customized

key (customizations results from the different possible

orders in this gene’s chain), The last base of the codon

selected (A,T,C or G) allows access to the bits 01, 10 ,00,

11 that form the message. It is important to note that the

marker gene is not altered by this operation. Information

on origin and dates of manufacture concerning these seeds

can be obtained. Security comes both the use of the key

and the specific nature of the changes, made voluntarily by

the manufacturer on the original gene, for obtain the de-

sired biological effects.

[6] M. Arita and Y. Ohashi, (2005) Secret signature inside

genomic DNA biotechnol. Prog, 20 (5), 1605-1607.

[7] K.Tanaka, A. Okamoto, and I. Saito, (2005) Public-key

system using DNA as a one way function for key distri-

bution. Biosystems, 81(1), 25-29.

[8] N. Yachie, K. Sekiyma, J. Sugahara, Y. Ohashi, and M. To-

mita, (2007) Alignment-based approach for durable data

storage into living organisms biotechnol. Prog., 23, 501-505.

[9] D. Heider and A. Barnekow, (2007) DNA-based water-

marks using the DNA-Crypt algorithm, BMC Bioinfor-

matica, 8, 176.

[10] W. L. Nison, N. Munakata, G. Horneck, H. J. Melosh,

and P. Setlow, (2000) Resistance of Bacillus endospores

to extreme terrestrial and extraterrestrial environments.

Micbiol. Mol. Bio. Rev., 64(3), 548-57.

Openly accessible at