They have been named Lokiarchaeota, partly after the Loki's Castle volcanic vent system lying 15km away from the site where the microbes' genetic material was isolated in cold marine sediments of the Arctic Mid-Ocean Ridge.
Domain names
Prokaryotes are single-celled organisms and comprise all bacteria and archaea (a group of microbes that were once considered to be bacteria, but now form a separate domain of life).
Eukaryotes comprise some single-celled life forms, as well as all multi-cellular organisms, such as animals, plants and fungi.
The gulf between these two groupings is vast. The eukaryotes possess cellular structures that are enclosed within a lipid envelope. The defining trait is the nucleus, which hosts the cell's genetic material.
Another would be mitochondria; these are found in most eukaryotes and act as a cell's "batteries". According to a widely accepted theory, mitochondria began as bacteria and were gradually incorporated into eukaryotic cells, perhaps through some relationship of mutual benefit.
Lokiarchaeota have genes that code for proteins only otherwise found in eukaryotes, such as parts of the cytoskeleton - a matrix that supports cell shape and movement.
Sister act
"Archaea and eukaryotes are sister groups, sharing a common ancestor," said lead author Thijs Ettema, from Uppsala University in Sweden.
He told BBC News: "This has been a leading model for 20 years or so. What happened a few years ago is that the branch in the tree that had the eukaryotes jumped on to the archaea branch. More specifically, it was affiliating with a group known as the TACK archaea."
Lokiarchaeota fall within the TACK grouping and represent the closest prokaryotic organisms to the eukaryote state.
According to Dr Ettema, the similarities between them show that Lokiarchaeota shared a common ancestor with eukaryotes roughly two billion years ago, and that this ancestor possessed a "starter kit" of genes that supported the increase in cellular complexity seen in eukaryotes today.
He explained: "The fact that we have found these same genes in [Lokiarchaeota] does not mean that they have the same function as they do in eukaryotes.
"But what we need to do to find out what those genes do in Lokiarchaeota is to carry out experiments, and for that we need actual cells."
The team had to reconstruct the new organisms from genetic material found in the cold marine sediments. But the effort to isolate cells will be a challenge.
"Getting the samples is not easy, and the amount of nutrients in these harsh environments is extremely limited. So the number of cells in these sediments will be extremely low and in general life down there is very slow.
"Some people have made predictions about how often cells divide down there and they have come up with numbers like one division every 10 years. If you want to grow them in the lab, these are not timescales that are feasible."
But the researchers are looking for "Loki-like" organisms in other locations, including hot springs in Yellowstone National Park, in the US, and New Zealand.
"We might even find Loki-like organisms that have more recent ancestry with eukaryotes. We could try to reconstruct their genomes and find additional pieces of the puzzle of how complex life might have originated," said Dr Ettema.
A key event in the evolution of eukaryotes was the acquisition of mitochondria. Lokiarchaeota do not possess them - making this organism no different from any other prokaryote. So precisely when cells first merged with the ancestors of these cellular powerhouses remains an open question.
"The acquisition of mitochondria really got things started," said Dr Ettema, adding: "The genes we find in Loki provide some pointers."
One critically important gene in eukaryotes is that which encodes a protein called actin. This has many functions in eukaryotic cells, but one of them is "phagocytosis". This process enables cells to engulf other cells, "eating" them.
"In Loki we also find genes that are related to those that encode actin proteins. Although we don't know what they do in Loki, we can infer that the last common ancestor had these genes," said Thijs Ettema.
Commenting on the research in the latest edition of Nature, Newcastle University cell biologists Martin Embley and Tom Williams write: "The identification of Lokiarchaeota so early in the history of this nascent field suggests that more closely related archaeal relatives of eukaryotes will soon be discovered."
A study by Facebook's own researchers has investigated whether the site's "news feed" filters content that users disagree with politically.
It finds that Facebook's algorithms do sift out some challenging items - but this has a smaller effect than our own decisions to click (or not) on links.
By far the biggest hit to ideologically "cross-cutting" content comes from the selection of our Facebook friends.
Other experts welcomed the study but also called for more, broader research.
Published in the journal Science, the new study was motivated by the much-debated idea that getting our news via online social networks can isolate us from differing opinions.
Critics have argued that this effect, particularly if exacerbated by "social algorithms" that select clickable content for us, is bad for democracy and public debate.
"People are increasingly turning to their social networks for news and information," said co-author Solomon Messing, a data scientist at Facebook.
"We wanted to quantify the extent to which people are sharing ideologically diverse news content - and the extent to which people actually encounter and read it in social media."
Choices, choices
Dr Messing and his colleagues studied 10.1 million US Facebook users, drawn from the 9% of adults on the site who declare their political affiliation as part of their profile - labelling themselves "liberal" or "conservative", for example.
The researchers found 226,000 news stories that were shared by more than 20 of those users, and gave a political "alignment" score to each one based on the stated ideology of the people who shared it.
Then they set about evaluating how much "cross-cutting" content, from the opposing political sphere, was available to each user.
If people accessed random posts from right across Facebook, the study reports that 40-45% of what they saw would fall into that category. But, of course, this is not how the Facebook news feed works, because we only see what our friends share.
Based on what their friends posted on the site, the users in the study could only ever have seen a news feed with 29.5% cross-cutting material.
This is quite a drop, largely produced by our tendency to "friend" people similar to ourselves; the study reported that on average, about 80% of a user's Facebook friends, if they declare a preference, have similar political views.
After we select our own friends, Facebook also refines what we see in the news feed using its complicated, controversial - and confidential - social algorithms. According to the study, this process did cut the proportion of cross-cutting material down further, but only to 28.9%.
Then, even within what users were offered by their Facebook feed, the links from contrasting perspectives received fewer clicks: only 24.9% of the content people clicked or tapped was cross-cutting.
These percentages varied between the self-declared "conservative" and "liberal" Facebook populations, but the case made by the Facebook researchers is clear: "If you average over liberals and conservatives, you can very easily see that the choice of the articles that you select matters more than the news feed algorithm," Dr Messing told the BBC.
But he agreed that our choice of online friends makes the biggest difference - and that, based on this study, reading news through that filter cuts out a lot of ideologically challenging material, relative to what is shared across Facebook.
The big, unanswered question is whether this places us in any more of an echo chamber than reading news directly from our favourite major media brands, or other sources we find for ourselves.
Vigilance required
In a commentary also published in Science, Dr David Lazer from Northeastern University in Boston said this was an awkward comparison: "It is not possible to determine definitively whether Facebook encourages or hinders political discussion across partisan divides relative to a pre-Facebook world, because we do not have nearly the same quality or quantity of data for the pre-Facebook world."
He also said this research topic was a crucial one, and "continued vigilance" would be required.
"Facebook deserves great credit... for conducting this research in a public way. But there is a broader need for scientists to study these systems in a manner that is independent of the Facebooks of the world."
Prof Mason Porter, a researcher at Oxford University, UK, who has also studied social networks, expressed similar sentiments.
"This is something that it's really important to try to disentangle," he told the BBC. He said the next step would be to compare the findings with what happens in other online environments.
"How much is this true in other social networks? The main thing for studies like this is that you want to repeat it in different ways. I think it's going to inspire a lot more studies, which is really one of the best things a scientific paper can do," Prof Porter said.
Patrick Wolfe, a professor of statistics and computer science at University College London, said the study was interesting despite some limitations in the data.
"I think the claims are reasonably plausible; I don't think the conclusion is that surprising," Prof Wolfe told BBC News.
He commented that biases in how the subjects were selected might mean that the findings did not reflect the wider population. In particular, to get their 10 million users the researchers first excluded the 30% of US Facebook users who visit on fewer than four days each week - as well as the 91% of adult users who do not declare their political affiliation.
The authors agree their study is limited, but say these decisions were a matter of extracting something relevant from Facebook's vast reserves of data.
"We're interested in the population who are regular social media users," Dr Messing said. "For people who log on once a month, this question isn't really relevant."
And the numbers are still large: "We have millions of observations," he said. "The error bars are smaller than the points on the graph."
Virtual reality firm Oculus VR says its much anticipated headset, the Rift, will go on sale to consumers within the first three months of 2016.
Until now, the firm had only sold "developer" editions that were targeted at video games creators and did not represent the Rift's final design.
Facebook bought Oculus VR for $2bn (£1.3bn) in 2014, despite it not having released a commercial product.
The release date means that HTC's rival helmet should be the first available.
The Taiwanese manufacturer has promised to sell its own virtual reality headset, the Vive, to the public before the end of the year.
Oculus' launch date could theoretically coincide with that of PlayStation's Morpheus headset, which Sony has said will be released at some point during the first six months of 2016.
Unlike Morpheus, both the Rift and the Vive are designed to work with PCs rather than a video games console.
"All the hype around virtual reality grew up around Oculus over the last two years, and not to come to market before what is looking like a compelling proposition from HTC represents a moderate surrendering of the initiative," commented Piers Harding-Rolls, head of games at the IHS Technology consultancy.
"But I don't think we're talking about it missing out on huge sales volumes at what will be an early stage of demand for VR."
The virtual reality helmets will allow wearers to see both computer-generated and filmed images directly in front of their eyes.
Users will be able to change their view by moving their heads.
The idea is to give them a greater sense of immersion, helping owners feel as if they are actually within a video game or interactive film or documentary, rather than just watching it.
There was a previous attempt to popularise virtual reality technology in the 1980s and 1990s, but the limitations and high cost of the products limited demand.
However, after Oculus successfully raised $2.4m by crowd-funding in 2013 to make new kit, there has been growing interest in the idea.
Samsung already sells versions of Oculus' kit that use the screens and motion sensors of its Galaxy S6 and Galaxy Note handsets.
But the forthcoming launch of headsets designed to work with more powerful computers has raised expectations of what the associated software will be capable.
"We've got these higher-end devices coming to market, but what we haven't seen yet is fully-realised content," said Mr Harding-Rolls.
"We've seen technical demos or the occasional short level. But the E3 video games expo will be an opportunity for all the stakeholders involved to show off content."
E3 will take place in Los Angeles in June.
For now, potential buyers are being teased with a promise that the Rift has a "more natural fit" and an "improved tracking system" than earlier models shown off by Oculus.
"In the weeks ahead, we'll be revealing the details around hardware, software, input, and many of our unannounced made-for-VR games and experiences coming," the firm said on its blog.
"Virtual reality is going to transform gaming, film, entertainment, communication, and much more.
"E3 is just around the corner - this is only the beginning," it added.
Facebook's challenge is that HTC's Vive is being released as part of a tie-up with Valve, which is one of the leading video game developers in its own right, as well as being the owner of the best-selling PC games platform Steam.
Microsoft has yet to reveal whether it has virtual reality plans for the Xbox.
The firm has, however, shown off an augmented reality headset called the HoloLens, that it says will allow PC users to see graphics superimposed over views of the real world.