Millions of people pick up gonorrhea every year, but the bacteria that cause the disease (Neisseria gonorrheae) have picked up something in return. They carry a little bit of human DNA within their genomes. It seems that the microbe behind the clap is partly human.

The human side of N.gonorrheae is a ‘LINE-1 (L1) sequence’ – a short piece of DNA that can copy and paste itself into new locations in the human genome. It has no obvious function beyond making more copies of itself, but it is very good at that. There are around half a million L1 sequences in the human genome and together, they make up a fifth of our DNA. And one of these sequences managed to hop into N.gonorrheae.

Mark Anderson and Steven Seifert from Northwestern University discovered the out-of-place DNA because the full genomes of 14 strains of N.gonorrheae have been completed and are publicly available. Within this database, they found a small DNA fragment that’s almost a perfect match to a human L1. After sequencing many more strains of the bacterium, Anderson and Seifert found the rogue L1 in around one in nine of them, and in none of their close bacterial relatives.

N. gonorrheae can invade the cells of its host, but it can’t break into the nucleus where most of the DNA is stored. So how did it manage to smuggle in an L1? Anderson and Seifert think that the fateful event happened when an infected cell died and broke apart, exposing its own DNA to the bacterium living inside it. Even then, there’s no easy route into N. gonorrheae’s genome. L1 can hop around a genome but only if it has the right landing sites. A bacterial genome doesn’t provide any. How it got in is anyone’s guess, but Anderson and Seifert speculate that the bacterium could have glued a fragment that contained L1 into a broken chunk of its own genome.

This complicated chain of events could explain why transfers of DNA from humans to bacteria are so rare (indeed, this is the first example of such a swap). Other genetic exchanges are far more common. Bacteria can swap genes as readily as humans swap opinions and N.gonorrheae’s own genome is a melting-pot of genes from several species. This “horizontal gene transfer” is a great way of injecting rocket fuel into evolution. By trading genes, bacteria can gain new powers in a single bound, including both offensive and defensive abilities.

These swaps are so pervasive that a sixth of the genome of Escherichia coli – a common gut bacterium – is made up of borrowed genes. On rare occasions, bacteria also trade DNA with their hosts. In one famous case, a species called Wolbachia managed to insert its entire genome into that of the fruit fly it infects. It’s still there to this day, creating a genetic fusion of fly and microbe. Transfers in the opposite direction, from host to bacterium, are rarer.

It’s not clear if N.gonorrheae’s human loan is actually doing anything, but it certainly hasn’t changed very much since it first leapt into the bacterium. All of the L1s in all of the different strains are very similar, and still in the same place. Some evolutionary pressure could be stopping the fragment from changing, implying a possible use. But it’s not clear what that might be – after all, Anderson and Seifert couldn’t find any differences between the strains that have human L1 and those that don’t. Alternatively, the L1 sequence might have hopped across very recently and hasn’t had time to change. This could explain why only a ninth of the bacteria have it.

There is another obvious possibility: the human DNA could have come from a human. As geneticists handle their samples, bits of skin or hair can fall in, adding human DNA to an otherwise pristine sample. This is a big problem. In a different study, Mark Longo from the University of Connecticut found human sequences in over 450 other genomes, from bacteria to wheat to zebrafish.

Longo found these stray sequences by accident. He was originally searching the zebrafish genome for the remains of ancient viruses that embedded themselves in the genomes of our ancestors. These genetic fossils are found in a variety of different animals and they have similarities that reflect their shared history. But Longo found something different – short pieces of DNA called Alu elements that are unique to humans and other primates. These sequences weren’t just superficially similar. They exactly matched their human counterparts and they couldn’t possibly exist in a zebrafish. They were clearly contaminants.

Like L1s, Alu can move around the genome. But Longo ruled out the possibility that this mobile DNA had hopped from humans to other species, just as L1 had done into gonorrhea. He looked for, and found, stretches of DNA that flank Alu in the human genome and that can’t hop from place to place. These sequences hadn’t integrated themselves into the different genomes. They were like stray pieces from a different jigsaw puzzle that had ended up in the wrong box. At least a fifth of published genomes contain these vagrant pieces.

This could cause problems for scientists who study animal evolution. But the challenges and the stakes are even higher when human DNA contaminates human samples. This might seem odd, but consider that we’ll soon reach a point when individual people can have their genomes sequenced cheaply. Doctors could make medical decisions based on these results and they could do so wrongly if one person’s genes are contaminating another’s.

Meanwhile, Jonathan Eisen, who studies microbe genomes, says the contamination problem also vexes scientists who study species that can easily swap genes. “On the one hand, contamination has possibly contributed to mistaken claims of lateral gene transfer in the past,” he says. “However, filtering out all “weird” DNA will lead one to miss real LGT. In the end this is going to be very hard to prevent completely.”

But there are certainly steps that scientists can take. In Longo’s study, it’s telling that only the genomes of flu viruses, which are handled with extreme care, were free of contamination. Longo checked 172 flu genomes and couldn’t find a single trace of Alu.

Rachel O’Neill, who led Longo’s study, says that scientists can reduce the odds of contamination by handling samples with the greatest, “forensics-level” care. They can also run independent tests in different labs to check their conclusions. Eisen adds that the ideal way to weed out stray sequences is to truly finish genome sequences (many “full” genomes are actually patchy drafts) and to check and cross-check everything.

In the end, this is about alerting the scientific community to the scale of the problem and encouraging people to focus on quality control and validation, as well as technological speed. O’Neill says, “The most important thing is to be aware of the possibility.” Eisen agrees, citing the importance of “[educating] everyone about possible problems so that we do not see 1000 papers on weird lateral transfers over the next few years.”

So could the human DNA in gonorrhea just be one of these contaminants? It’s unlikely. Anderson and Seifert went back to the original strains whose genomes were sequenced and analysed them again. They got the same result, and the fact that the stray L1 appears at the same position in three different strains sealed the case. This seems to be one of those “weird lateral transfers” that’s actually genuine.

Reference: Anderson, M., & Seifert, H. (2011). Opportunity and Means: Horizontal Gene Transfer from the Human Host to a Bacterial Pathogen mBio, 2 (1) DOI: 10.1128/mBio.00005-11

Longo, M., O’Neill, M., & O’Neill, R. (2011). Abundant Human DNA Contamination Identified in Non-Primate Genome Databases PLoS ONE, 6 (2) DOI: 10.1371/journal.pone.0016410

More on horizontal gene transfer:

• Genes from Chagas parasite can transfer to humans and be passed on to children
• An entire bacterial genome discovered inside that of a fruit fly
• Jumping genes mobilise in the brains of people with Rett syndrome
• ‘Friendly’ genes are more likely to be passed around

February 16th, 2011 by Ed Yong in Bacteria, Genetics, Genomics, Horizontal gene transfer | 5 comments | RSS feed | Trackback >