🫘 Your Gut Bacteria Just Solved the Kidney Waitlist Crisis

Science, apparently, has a thing for stealing good ideas from places we'd rather not think about. This week: enzymes from your intestinal bacteria are converting kidneys into universal donor organs, Cambridge researchers are growing blood in dishes that literally turn red enough to see, and a 25-year-old meditation enthusiast is picking your future baby's IQ despite mathematical impossibilities in his models. Meanwhile, Type O patients are still dying on waitlists at 11 per day, which is either a policy failure or proof we needed gut microbes to save us all along.

Chinese and Canadian researchers successfully transplanted a Type A kidney into a Type O recipient after converting it to universal blood type using enzymes originally found in human stool samples. The kidney, treated with two bacterial enzymes during standard preservation, functioned for three days without the hyperacute rejection that normally destroys mismatched organs within minutes.

The enzymes come from Flavonifractor plautii, a gut bacterium that's been living in your intestines this whole time, quietly evolving to digest the exact same sugar molecules that define blood types. UBC's Dr. Stephen Withers discovered them in 2019 by screening human fecal DNA, because sometimes the solution to a $100 billion organ shortage crisis is literally sitting in the toilet. The enzymes work like molecular scissors, snipping off Type A antigens to reveal the universal Type O underneath. Treatment takes two hours. Just add to the perfusion fluid during transport, rinse, and transplant.

The study published in Nature Biomedical Engineering used a brain-dead recipient with high levels of anti-A antibodies (worst-case scenario) and gave NO immunosuppression drugs. The kidney removed 96.4% of A-antigens and functioned beautifully. By day three, antigens started regenerating (the body "remembers" its blood type), but single-cell sequencing revealed something unexpected: genes for immune tolerance were activating. The body was learning to accept the organ.

This matters because 52.8% of kidney waitlist patients are Type O, but they can only receive Type O organs. Meanwhile, Type O kidneys get allocated to anyone because they're universally compatible, leaving Type O patients waiting 2-4 years longer than other blood types. Eleven people die every day in the U.S. waiting for kidneys. If you can convert Type A organs (40% of the donor pool) to Type O, you nearly double the supply for the people who need it most.

Avivo Biomedical, the UBC spin-off commercializing this, is pursuing clinical trials. The challenge: antigens regenerate within days, so you'll need standard immunosuppression alongside enzyme treatment. But compared to traditional ABO-incompatible transplants (days of plasmapheresis, heavy drugs, living donors only), this is faster, cheaper, and works with deceased donors where every hour counts. Dr. Withers called it "years of basic science finally connecting to patient care." Turns out your microbiome has been carrying the answer all along. This is fine.

Science Corporation's PRIMA implant restored clinically meaningful vision to 81% of patients with advanced age-related macular degeneration in a pivotal trial, with some reading pages in books after a decade of blindness. The device is a 2mm photovoltaic chip, thinner than half a human hair, placed under the retina, powered wirelessly by infrared light from camera-equipped glasses.

The New England Journal of Medicine study followed 38 European patients for a year. Before implantation, they had "absolute scotoma" (complete black holes where central vision used to be). After: 84% could read letters, numbers, and words. Average improvement was five lines on an eye chart. One patient improved by 59 letters, which is basically getting someone else's eyes. Sheila Irvine, 65, went from "two black discs in my eyes" to reading prescription labels and doing crosswords. She called the moment she first saw a letter "dead exciting."

But how does it work? Your pocket processor runs AI algorithms that capture what you're looking at, the glasses beam it as pulsed infrared onto the subretinal implant, 378 photovoltaic pixels convert light to electrical current independently (no battery, no wires), signals stimulate surviving retinal neurons, and your brain interprets them as vision. The infrared is invisible to your remaining peripheral photoreceptors, so you use natural peripheral vision simultaneously with prosthetic central vision. It's a solar panel in your eye powered by invisible lasers, because of course it is.

Training takes months. Your brain needs to learn what these new signals mean, like cochlear implants. Currently black-and-white only (grayscale coming in software update). Resolution is limited, zoom helps (up to 8x magnification), you won't get 20/20 vision back. But for 5 million people worldwide with geographic atrophy (advanced dry AMD) who have zero treatment options to restore vision, being able to read food labels is life-changing.

Science Corp, led by Max Hodak (Neuralink's former president who left Elon's operation and is now actually getting working BCIs into humans), acquired PRIMA after the original French developer went bankrupt. CE mark application has also been submitted for European approval, expected 2026 availability. The FDA process is underway in the U.S., too. Cost is not yet announced, however, but Argus II (the discontinued predecessor with 60 pixels and crappy vision) was $150,000. PRIMA has 6x more pixels and delivers actual form vision, not just light perception. Next generation: 10,000 pixels, 20-micron resolution, potentially 20/20 with zoom. Stanford's Daniel Palanker conceived this in 2005, working on laser eye surgery, and thought, "wait, the eye is transparent, we should just beam light into it." Twenty years later: it works.

Nucleus Genomics launched Origin, an AI model claiming to predict human longevity from embryo DNA with unprecedented accuracy, trained on 1.5 million genomes and 7 million genetic markers. The 25-year-old founder says it enables "greater than 50% risk reduction" in diseases like Alzheimer's, diabetes, and cancers for IVF patients willing to pay $30,000. Meanwhile, an independent investigation found mathematical impossibilities in customer reports and accused the company of using outdated models incorrectly specified" to appear more accurate than they are.

Competitor Herasight's analysis of archived Nucleus code found some disease models based on only 12 SNPs (laughably few), customer reports showing "impossibly good results," and validity "far beyond any competitors or academic results." For example, one embryo with an elevated ADHD polygenic score but below-average estimated risk, which is mathematically incoherent. Herasight explicitly excluded Nucleus from their competitive benchmark because including them "would not yield a reliable or meaningful addition to our analysis." That's the scientific equivalent of "we don't want whatever they're smoking."

Nucleus offers predictions for nine age-related diseases plus cognitive ability, height, eye color, hair color, and other traits. The models are "open-weighted," meaning researchers can download and inspect them, which is how the problems got discovered. The company operates as a Laboratory Developed Test (LDT) under CLIA certification, sidestepping FDA approval requirements. Founder Kian Sadeghi defended the practice by saying parents have the liberty to choose blue eyes over green eyes for their babies, which is either progressive reproductive freedom or eugenics with venture capital, depending on who you ask.

The broader scientific consensus is also harsh. The American College of Medical Genetics concluded polygenic embryo screening "has not been proven to provide clinical utility." A Nature article called it "an unproven, unethical practice" in 2021. The fundamental problem: polygenic scores work reasonably well at the population level but fall apart for within-family predictions (comparing genetic siblings). Environmental factors, pleiotropy, regression to the mean, and limited embryo numbers (typically 1-10 per IVF cycle) all reduce effectiveness. Absolute risk reduction is often 0.12% to 8.5% even when relative risk reduction looks impressive.

Meanwhile, Nucleus is already "in use with IVF+ partners" in Los Angeles, New York, and Mexico City. No peer-reviewed publications validate Origin's specific claims. Peter Thiel's Founders Fund invested $32 million in total with other investors. The company frames this as giving children "more than they had" while critics see genetic class divisions and designer baby dystopia. The real villain? Regulatory gaps that let companies sell accuracy claims without proving them, leaving consumers to discover through competitor investigations that the math literally doesn't add up.

Cambridge researchers created hematoids, self-organizing clusters of stem cells that spontaneously develop into miniature blood-producing factories complete with beating heart cells to circulate fluid. By day 13, the structures turn visibly red with hemoglobin, producing the type of advanced blood stem cells needed for therapeutic use. It's a microscopic circulatory system in a dish, which is either regenerative medicine's future or proof that scientists have completely lost the plot.

The breakthrough, published in Cell Reports, shows these "embryo-like" structures replicate weeks 4-5 of human development (the hidden period after implantation when you can't observe real embryos). Start with pluripotent stem cells, add one TGF-β inhibitor, and stand back. By day 2, three germ layers form. Day 8: heart cells start beating. Day 13: blood appears. No complex growth factor cocktails, no forced gene expression, just cells self-organizing through their intrinsic developmental program. Dr. Jitesh Neupane reported, "It was an exciting moment when the blood red colour appeared in the dish - it was visible even to the naked eye." What makes hematoids special is that they produce definitive hematopoietic stem cells (the advanced "second wave" of blood development) that can become T cells, NK cells, red blood cells, and multiple lineages. Previous methods required expensive cocktails of exogenous cytokines. This uses self-generated signals from the structures themselves, including liver-like hepatocytes that appear spontaneously. The beating hearts aren't just decorative; they create circulation for development. Interesting Engineering: It's the "no assembly required" approach to blood production.

Could this solve blood shortages? Blood donations dropped 40% over 20 years globally, only 3% of eligible Americans donate annually, and the Red Cross reports a critically low national inventory. Someone needs blood every two seconds in the U.S. Type O supplies are dangerously low. Hematoids are "still in early stages" with no clinical timeline. There are challenges, of course. Scaling production, improving stem cell maturation, extensive safety testing, regulatory approval, and cost-effectiveness. Realistically, many years to decades before you're getting a hematoid-derived transfusion.

The University of Cambridge team patented the technology through Cambridge Enterprise and received ethics approval under the UK's Code of Practice for stem-cell-based embryo models. Important: hematoids deliberately lack the tissues needed to develop into actual embryos (no yolk sac, no placenta). They're engineered to be useful but not viable. Still, using "embryo-like" structures for blood production raises philosophical questions about where science ends and creation begins, though regulatory bodies deemed this acceptable. For now, it's proof-of-concept that your own skin cells could theoretically be turned into perfectly matched blood for transfusions. Whether that's regenerative medicine's holy grail or an expensive solution to a donation problem depends on how optimistic you're feeling.

RESEARCH & NEWS

Counterintelligence veterans are warning that Chinese and Russian intelligence agencies are deploying attractive operatives to seduce tech workers and investors for IP theft, using tactics like "love bombing," fake social media validation, and emotional manipulation. We can see some evidence around. A mix of documented prosecutions for economic espionage (Genentech trade secrets stolen for $1 billion Taiwanese biotech, DuPont formula theft), and anecdotal reports of LinkedIn honeypots. Actual honey trap cases with hard evidence are rare; most claims come from anonymous sources, and the most dramatic allegations (operatives marrying targets and having children for decades-long operations) lack verified examples.

James Mulvenon, chief intelligence officer at Pamir Consulting, reports "an enormous number of very sophisticated LinkedIn requests from the same type of attractive young Chinese woman" that "really seems to have ramped up recently." At a Virginia conference on Chinese investment risks, two attractive Chinese women attempted entry with detailed participant information and were denied. Meanwhile, former Russian spy Aliia Roza (turned whistleblower and Beverly Hills influencer) describes training from age 18 as a "master manipulator" using techniques like appearing in targets' lives seven times before making contact, manufacturing social connections, and hero-complex exploitation. Her credibility, though unverified, lacks independent confirmation of an intelligence background, but it makes for excellent podcast content.

What's definitely real is $600 billion annual losses to IP theft, primarily in China. FBI Director Christopher Wray calls Chinese espionage "the greatest security threat" facing the U.S. The Bay Area handles 70% of national economic espionage cases. Documented prosecutions include Jerry Chun Shing Lee (former CIA officer, 19 years for helping dismantle the CIA's China informant network), Walter Lian-Heen Liew (15 years for stealing DuPont titanium dioxide formula), and JHL Biotech executives who pleaded guilty to a decade-long theft of Genentech trade secrets to build a "nearly $1 billion Taiwanese unicorn built on foundation of lies." That last one is biotech-relevant and absolutely confirmed.

The playbook, per Roza's descriptions: target single tech workers with access to sensitive information, create emotional dependency through floods of attention, isolate them from colleagues ("your boss doesn't appreciate you"), extract secrets under pressure ("if you don't send this right now, I'll disappear forever"), then either blackmail with recorded encounters or vanish. Targets are "brilliant in technology but fragile in relationships." The real asymmetric advantage: The U.S. doesn't employ similar tactics legally or culturally. Elon Musk's response to this is humorous too: "If she's a 10, you're an asset." Which is either peak internet humor or deeply unhelpful, depending on your perspective.

The problem is distinguishing legitimate threat (verified prosecutions, FBI warnings, billions in documented losses) from sensationalized "sex warfare" narrative (limited evidence, anonymous sources, clickbait angles). Intelligence agencies definitely target Silicon Valley. Biotech is absolutely in the crosshairs (COVID vaccine espionage, GenScript investigation, repeated cyberattacks). Sexual entrapment is a historical espionage tactic (Cold War Romeo spies, Anna Chapman). Current evidence of widespread honey trapping in tech? Thin. Mostly consists of "my colleague knows a guy" stories and LinkedIn profile observations. Futurism correctly noted, reporting "not always clear whether accusations are grounded in fact or more of a hunch." To be fair, that's kind of how counterintelligence works, but it makes writing a fact-checked newsletter complicated. What’s the takeaway? Industrial espionage is massive, biotech is targeted, exercise caution with attractive strangers offering investment advice, and maybe don't trust that your gut bacteria are the only microscopic threat you're carrying.

We refuse to pretend science isn't absurd while taking the implications deadly seriously. We've covered gut bacteria solving organ shortages, retinal implants restoring vision with pocket computers, AI embryo selection with impossible math, blood factories with tiny hearts, and the blurry line between legitimate espionage concerns and sensationalized spy drama.

Biotechnology is accelerating faster than our ability to verify accuracy claims or regulate ethical boundaries, but the breakthroughs that survive scrutiny (PRIMA's 80% success rate, enzyme-converted kidneys functioning in humans, hematoids producing real blood) are something transformative. The ones that don't (embryo selection models with mathematical impossibilities) reveal how easily venture capital, founder charisma, and regulatory gaps can replace peer review with marketing.

What are you most excited or terrified about? Hit reply. Share your expertise if we missed something. Forward this to that friend who still thinks science is boring (how?). And if you're Type O and waiting for a kidney: hold on. Your microbiome might have already solved it.

Keep questioning everything (especially the math),
Prateek & Jere

P.S. The Cambridge researchers who created hematoids patented them through the university's innovation arm. The enzymes that convert kidneys were discovered by screening human fecal samples. The most elegant solutions increasingly come from places we'd rather not look, which tells you something about how innovation actually works versus how we wish it worked.