Prime Edit Zero

• Chapter 1 The Blueprint
• Chapter 2 Anomalous Base Pairs
• Chapter 3 Proprietary Sequence
• Chapter 4 The First Anomaly
• Chapter 5 Corporate Patent
• Chapter 6 The Silent Gene
• Chapter 7 Ghost in the Helix
• Chapter 8 Synthetic Triggers
• Chapter 9 The Architect's Warning
• Chapter 10 Code Redundancy
• Chapter 11 The Ticking Strand
• Chapter 12 Breach of Contract
• Chapter 13 In Vivo
• Chapter 14 The Prime Registry
• Chapter 15 Cellular Decay
• Chapter 16 Off-Target Effects
• Chapter 17 Black-Market Sequencing
• Chapter 18 The Override Protocol
• Chapter 19 Forced Termination
• Chapter 20 Epigenetic War
• Chapter 21 The Lab in the Shadows
• Chapter 22 Rewriting the Switch
• Chapter 23 System Purge
• Chapter 24 Zero State
• Chapter 25 Expression
• Chapter 26 Unbound

The human genome contains roughly three billion base pairs, which sounds like a lot until you realize it can fit comfortably on a cheap thumb drive. If you compress it properly, you can carry the entire structural design of a person in your coin pocket right next to a crumpled coffee shop receipt.

Eden Vance knew this because she had carried her own blueprint around for three years. It sat inside a brushed-aluminum drive attached to her keychain, clanking softly against her apartment key every time she turned a lock.

To the rest of the world, Eden was an answer to a late-twentieth-century bioethics exam question. She was Prime Edit Zero: the first human being successfully brought to term whose genome had been altered not by crude viral vectors or the sledgehammer precision of early CRISPR, but by second-generation prime editing. No double-stranded DNA breaks. No chaotic insertions or off-target deletions. Just a surgically precise search-and-replace command written directly into the nucleus of a single-cell zygote twenty-four years ago.

Her parents had been carriers for Huntington’s disease, both possessing the treacherous, repeating CAG expansion on chromosome 4 that guaranteed a slow, neurological disassembly in middle age. Under the older genetic regime, their options were limited to selective embryo discard or gambling on pure chance. Under Nexus Biome—the multinational life-sciences conglomerate that held the exclusive patents on reverse-transcriptase editing complexes—they were offered a third option.

They got Eden. Perfect, neuro-typical, disease-free Eden.

"You look remarkably unbroken for a historical monument," Dr. Marcus Hayes said, lowering his tablet. He was sitting on a swivel stool across from her in the diagnostic suite of the Nexus Metropolitan Clinic. The room smelled of ozone, isopropyl alcohol, and the faintly sweet aerosol fragrance they sprayed to keep patients from realizing they were sitting inside a multi-billion-dollar corporate asset management facility.

Eden adjusted her sleeve, pulling it down over the small red puncture mark in the crook of her elbow. "I feel uncomfortably normal, Marcus. It’s the media that keeps insisting I should be glowing in the dark."

"Give it time," Hayes smirked. He tapped the screen of his interface, sending her latest blood panel into the cloud. "Your metabolic markers are baseline. Liver enzymes are clean. Your telomeres look like those of a fresh-faced teenager, which is technically unfair to the rest of us suffering through natural cellular senescence."

"And the sequence?" Eden asked. She tried to keep her voice light, as if asking about the tread wear on a used sedan.

Hayes paused, his thumb hovering over the screen. The flickering fluorescent light caught the silver frames of his glasses. "The sequence is what it’s always been. Clean. Exactly as programmed by the founding committee in 2028. You are, quite literally, the most thoroughly verified biological organism on the eastern seaboard."

"Can I get the raw FASTQ files this time?"

Hayes looked up, his brow furrowing slightly. "Why? You’re an environmental data analyst, Eden. You process climate models and urban runoff charts. What do you want with three gigabytes of raw adenine, cytosine, guanine, and thymine?"

"It’s my blood," Eden said simply. "And I bought a new workstation last month. I want to render my own three-dimensional chromatin folds. Just a hobby project."

It wasn’t entirely a lie. She did have a high-performance workstation in her living room, and she did spend her evenings writing scripts to parse large environmental datasets. But the desire to look at her own code wasn't driven by idle curiosity. For the past six months, Nexus Biome had been quietly updating its patient end-user agreements. The notifications had arrived in her inbox like any other digital clutter—wordy, legalistic notices regarding "updated terms of service for proprietary genetic constructs."

Most people didn't read terms of service for their phones, let alone their cellular physiology. But Eden had a vestigial habit of reading fine print, inherited from her mother, a contract lawyer who had spent the last decade of her life ensuring Nexus didn't own the copyright to Eden’s facial features.

"I’ll have to request clearance from legal to release the unmasked raw sequence," Hayes said, though his hand was already moving across the screen to approve the transfer. He had known Eden since she was six years old; he had been the young research assistant holding the clipboard while senior geneticists checked her reflexes and measured her skull growth. He possessed the mild, casual affection of a mechanic who had serviced the same high-end sports car for two decades. "They’re always skittish about sequence leaks. Proprietary enzymes, competitive advantage, all that nonsense."

"It’s my DNA," Eden reminded him gentle.

"Legally? It's a joint venture," Hayes chuckled, though there was no humor in his eyes. "You supply the carbon, they supplied the proofreading. But fine. Check your secure portal tonight. The raw reads will be there. Just don't go selling your polymerase maps to the Chinese."

The rain was falling in a steady, lukewarm drizzle by the time Eden left the clinic. The Nexus Biome tower towered fifty stories over the center of the city, a sleek monolith of black glass and white steel shaped vaguely like a stylized nucleotide strand. At night, the lighting along its facade pulsed in a rhythmic, slow wave that mirrored a resting heart rate.

She walked three blocks to the light rail station, her head down against the damp wind. Passersby didn't recognize her, and she preferred it that way. When she was twelve, there had been a brief, terrifying window of media obsession—documentaries about "The Prime Generation," photo shoots where she was made to hold potted plants and look thoughtfully into the camera as if she were the patron saint of agriculture. But the world had moved on. Gene editing had become faster, cheaper, and far less novel. Today, thousands of children were born with edited traits, though none carried the specific, historical weight of being Zero.

When she reached her apartment—a narrow, two-bedroom unit on the fourth floor of a converted brick warehouse—she didn't bother turning on the main lights. She kicked off her boots, brewed a pot of black tea, and booted up her desktop system.

The screen illuminated her face in a cool blue glow. A notification flickered in the corner of her display: 1 New Secure Transfer - Nexus Biome Legal & Diagnostic Portal.

She downloaded the archive file. It was massive: a raw, unaligned genomic dataset generated by a state-of-the-art nanopore sequencer. Three billion pairs of molecular information, represented by four letters repeated in seemingly endless, chaotic combinations.

Eden opened a terminal window and loaded her personal analysis environment—a suite of open-source bioinformatics tools she had customized over the weekend.

"Let's see the blueprint," she muttered to herself, taking a sip of tea.

In the early days of prime editing, the process was described to the public using simple, comforting metaphors. The editors were "word processors," the guide RNAs were "search terms," and the reverse transcriptase was the "keyboard" that typed the correct sequence into the living cell.

What they didn't tell the public was that word processors leave metadata. Every time a digital document is edited, hidden layers of information track who opened the file, what was changed, and when it was saved.

Eden’s target was Chromosome 4.

She brought up the reference model for the human genome—the standard, wild-type sequence that served as the baseline for the human species—and aligned her personal dataset against it. Her screen split into two horizontal windows. On the top was the standard human reference. On the bottom was Eden Vance.

The scroll bar sped past millions of letters. Most of it matched perfectly. A few natural single-nucleotide polymorphisms appeared here and there—harmless variations that made her height slightly above average and gave her eyes a pale, hazel tint.

Then she hit the targeted site: position 3,076,604 on Chromosome 4.

In a natural carrying individual with Huntington’s, this region contained an abnormal run of CAG repeats—sometimes forty, fifty, or sixty in a row—coding for a toxic, misfolded protein that slowly destroyed the striatum of the brain.

In Eden’s sequence, the CAG run was precisely nineteen repeats long. Healthy. Normal. Stable.

She zoomed in further, looking at the flanking regions—the stretches of non-coding DNA that sat on either side of the edited gene. This was where the prime editor’s pegRNA would have docked two decades ago.

She expected to see the standard scarless transition promised by Nexus Biome’s old press releases. Zero trace architecture, the brochures had claimed back in 2028. A smooth, natural genome with no synthetic signatures left behind.

Instead, she saw something else.

Immediately following the edited codon sequence, where the intron usually resumed its long, quiet stretch of junk DNA, there was a block of base pairs that didn't match the human reference genome. Nor did it match any known biological sequence in the National Center for Biotechnology Information database.

It was roughly two hundred base pairs long. It wasn't a random insertion—random insertions were chaotic, filled with stuttered repeats and broken reading frames. This block was crisp. The distribution of adenine, cytosine, guanine, and thymine was balanced, structured, and unmistakably artificial.

Eden stared at the screen, her tea growing cold in her hands.

"That shouldn't be there," she whispered to the empty room.

She ran a basic statistical check on the sequence segment. The GC-content—the percentage of nitrogenous bases that were either guanine or cytosine—was exactly fifty percent. In nature, non-coding junk DNA flutters wildly in its composition. Fifty percent on the nose was the mark of synthetic design.

She pulled up her terminal and ran a simple translation script, converting the base pairs into amino acid sequences to see if it coded for a protein. Nothing. The reading frame was littered with stop codons. It was a genetic dead end. It didn't make anything. It just sat there, silent, anchored to the backbone of her fourth chromosome like a tiny piece of unread text left in the margins of a master manuscript.

She opened a text editor, isolated the strange two-hundred-base-pair sequence, and displayed it in plain text across her screen:

GCTAGCTAGCTAGCTAGCTAGCTAGCTAGCTAGCTAGCTAGCTAGCTAGCTAGCTAGCTAGCTAGC

Except it wasn't a simple repeating pattern like that. As she stared at the actual letters, her mind, trained to spot subtle anomalies in atmospheric pollution matrices, recognized a different kind of structure.

It wasn't a protein code. It was data.

Eden sat back in her chair, her heart beginning to beat a fraction faster against her ribs. She looked at her thumb drive sitting on the desk beside her keyboard—the shiny aluminum casing that carried her baseline files. She had assumed for years that her blueprint was a clean, simple correction of a fatal flaw. A gentle touch-up on a master painting.

She had spent her entire life thinking she was just a human who had been saved by technology.

She looked back at the screen, where the synthetic block of DNA glowed in monochromatic white text against the black terminal window. It sat inside every single one of her fifty trillion cells. It had been there when she learned to walk, when she graduated from school, when she had her first kiss, and when she buried her mother.

It wasn't a repair. It was a stamp.

Eden placed her fingers on the keyboard, opened a fresh Python script, and began to write a program to crack the encoding of her own blood.

The human brain is remarkably good at convincing itself that an impossible thing is merely a software glitch.

Eden spent twenty minutes staring at the screen, convincing herself that the extra two hundred base pairs were a computational artifact. Sequencers made mistakes all the time. Nanopore sequencing—where single strands of DNA are pulled through microscopic proteins while an electrical current measures fluctuations in the ionic flow—was notoriously noisy. A transient voltage spike caused by a microscopic bubble or a stray lipid molecule could easily be interpreted by the base-calling algorithm as a string of phantom nucleotides.

She opened the raw signal telemetry file. The nanopore raw data, stored in a massive binary format, preserved the exact picoampere readings recorded as her DNA translocated through the synthetic pore.

She pulled up the signal trace for position 3,076,604 on Chromosome 4.

The ionic current graph was razor-sharp. There were no voltage drops, no thermal noise spikes, no signal stuttering. The signal levels for the anomalous region stepped cleanly between four distinct, unambiguous electrical plateaus. Each plateau corresponded to a specific nitrogenous base: adenine at 12 picoamps, cytosine at 18, guanine at 24, and thymine at 30.

The sequencer hadn't hallucinated. The base caller hadn't stumbled. The bases were real, physical molecules of phosphate, sugar, and nitrogen, linked together by covalent phosphodiester bonds inside her cell nuclei.

Eden leaned back in her mesh ergonomic chair and let out a long, slow breath that rattled against the glass of her half-empty tea cup.

"Okay," she said aloud to the quiet room. "It’s real. It’s physical. Now what are you?"

She started with standard bioinformatic sanity checks. In genetics, anomalous insertions were usually one of three things: transposons—so-called "jumping genes" that naturally copy and paste themselves across the genome; viral integrations, left behind by ancient retroviruses that had infected human ancestors millions of years ago; or structural variants caused by cosmic radiation or replication error.

She ran a BLAST alignment—the Basic Local Alignment Search Tool used by biologists worldwide to search the GenBank repository of known biological sequences. She fed her two-hundred-base-pair snippet into the query window and hit execute.

The search swept through billions of sequences: bacterial plasmids, fungal chromosomes, mammalian genomes, synthetic cloning vectors, plant plastids, and viral envelopes.

Ten seconds later, the screen refreshed with zero results.

No match in Homo sapiens. No match in Escherichia coli. No match in Saccharomyces cerevisiae. No match in any known synthetic plasmid used in academic research.

The sequence was an orphan.

Eden rubbed her eyes. The clock on her desktop monitor read 1:14 AM. Rain continued to tap rhythmic, irregular patterns against her windowpane, throwing streaks of yellow light from the streetlamp below across her ceiling.

She opened a terminal window and began writing a script to evaluate the information density of the sequence. Biological DNA, even non-coding intron regions, has a specific entropy signature. Natural selection creates patterns: short repeating motifs, regulatory binding sites, microRNA targets. Purely random synthetic noise has high Shannon entropy, while biological code is slightly more predictable.

She wrote a quick script in Python using the standard scientific library to compute the Shannon entropy of her sequence compared to the surrounding natural DNA:

import math
from collections import Counter

def calculate_entropy(sequence):
    counts = Counter(sequence)
    length = len(sequence)
    entropy = 0.0
    for base, count in counts.items():
        freq = count / length
        entropy -= freq * math.log2(freq)
    return entropy

seq_anom = "GCATCGATCGATCGATCGATCGA..." # 200 bp insert
seq_flank = "AAATGCATTACTAGCTAGCTAGC..." # 200 bp flanking human intron

print("Anomalous Entropy:", calculate_entropy(seq_anom))
print("Flanking Entropy:", calculate_entropy(seq_flank))

She hit run.

The output flashed in yellow text at the bottom of her terminal:

Anomalous Entropy: 1.9982 bits per base
Flanking Entropy: 1.7412 bits per base

The theoretical maximum entropy for a sequence of four bases (A, C, G, T) is exactly 2.0 bits per base—meaning every single base carries two bits of information with zero redundancy. Natural human DNA sits around 1.7 to 1.8 bits per base due to evolutionary repeats and structural constraints.

Her anomalous insert sat at 1.9982.

It was almost pure, unadulterated information. It had been optimized for maximum data density, stripped of all biological padding.

"It's an encoded file," she whispered.

In the mid-2020s, before she was born, bioengineers had started experimenting with DNA-based data storage. Because DNA is millions of times denser than silicon flash memory and can endure for thousands of years if kept dry and cold, tech companies had developed simple mapping schemes to write digital data into synthetic DNA strands.

The simplest scheme was quaternary-to-binary mapping: adenine and cytosine represented 00 and 01, while guanine and thymine represented 10 and 11. Other schemes mapped three-base codons directly to ASCII text characters or base64 strings.

Eden pulled up a fresh scratchpad in her editor and wrote a translation script to test the standard mapping schemes.

If you map four bases to two bits each, two hundred base pairs equals four hundred bits, or fifty bytes. Fifty bytes wasn't enough for a high-resolution image or an executable binary, but it was plenty of space for a string of text, a serial number, or a cryptographic hash.

She set up her mapping dictionary:

# Quaternary encoding map
mapping = {'A': '00', 'C': '01', 'G': '10', 'T': '11'}

def dna_to_bytes(dna_str):
    bit_string = ''.join(mapping[base] for base in dna_str)
    byte_array = bytearray()
    for i in range(0, len(bit_string), 8):
        byte = bit_string[i:i+8]
        byte_array.append(int(byte, 2))
    return byte_array

She ran the script against the two-hundred-base-pair string.

The screen printed fifty hex bytes:

50 52 4F 50 52 49 45 54 41 52 59 3A 20 4E 45 58 55 53 2D 42 49 4F 4D 45 2D 50 41 54 2D 32 30 32 38 2D 30 30 31 2D 5A 45 52 4F 00 00 00 00 00 00 00 00

Eden stared at the output, her pulse hammering against her throat.

She knew those hex codes. In computer science, 50 52 4F 50 were the ASCII values for capital letters.

She added an ASCII decoding line to her script: print(bytes.decode('ascii', errors='ignore')) and pressed Enter.

The plaintext message rendered across her terminal in high-contrast white text:

PROPRIETARY: NEXUS-BIOME-PAT-2028-001-ZERO

The remaining bytes were padded with trailing null characters, filling out the rest of the two-hundred-base-pair block.

Eden sat frozen.

It wasn't a biological artifact. It wasn't an accidental insertion left behind by a messy reverse transcriptase complex. It was a digital copyright tag. A cryptographic watermark burned into position 3,076,604 on Chromosome 4, replicated across every single nucleus in her brain, her heart, her skin, and her lungs.

Nexus Biome hadn't just fixed her Huntington's mutation. They had stamped their corporate emblem onto her genome like a brand on livestock.

By 3:30 AM, the rain had stopped, leaving behind a cold, thick fog that swallowed the lower floors of the surrounding brick buildings. Eden hadn't moved from her chair.

Her mind was racing through the implications.

When Nexus Biome developed the prime editing treatment for her parents in 2028, the company had been locked in a fierce, multi-front patent war with rival biotechnology firms in Boston and Zurich. Millions of dollars had hinged on proving who owned the specific pegRNA designs and molecular complexes used in human trials.

A watermarked sequence made sense from a legal perspective. If a competitor stole their edited cells or tried to clone their proprietary lines, Nexus could sequence the target region, find the string PROPRIETARY: NEXUS-BIOME-PAT-2028-001-ZERO, and immediately establish ownership in a federal court.

It was arrogant, invasive, and deeply unethical—but it made cold corporate sense.

Except for one detail.

A pure data tag sitting inside an intron was supposed to be silent. Introns are spliced out during pre-mRNA processing; they are treated as genomic scaffolding, ignored by the ribosomes that build proteins. A tag sitting in an intron should just float there, inert and invisible to the cell's machinery.

Eden pulled up her local copy of the full FASTQ read set again. She decided to check the quality scores of the anomalous region across multiple independent sequencing reads.

Modern nanopore machines didn't just read a strand once; they read thousands of overlapping fragments, aligning them to build a high-confidence consensus sequence.

She parsed through the individual read alignments that covered position 3,076,604.

Most of the reads showed the watermark sequence she had just decoded. Clean, clear, uniform.

But as she scrolled down to the reads with higher coverage depth—reads that captured longer stretches of DNA extending further down Chromosome 4—she noticed a secondary pattern.

The watermark wasn't an isolated island.

Roughly five hundred base pairs downstream from the watermark, still well within the non-coding space of Chromosome 4, sat another set of non-human base pairs.

Eden felt a sudden cold sensation in the pit of her stomach.

She grabbed her mouse, her hand trembling slightly, and selected the second block of anomalous code. It was longer than the watermark—nearly eight hundred base pairs.

She ran her Shannon entropy script against the new block.

Entropy: 1.9995 bits per base

Another artificial data block.

She fed it directly into her ASCII decoder, expecting another legal string, perhaps a full patent claim or a corporate address.

The script executed. The output displayed:

[ERROR: Non-ASCII characters detected. Binary payload.]

The second block wasn't text. It was compressed binary data.

Eden stood up from her desk so fast her chair rolled backward and bumped against her bookshelf, knocking over a stack of paperback novels. She walked over to the kitchen counter, turned on the tap, and splashed cold water onto her face.

Her reflection in the small bathroom mirror looked pale, her hazel eyes wide and shadowed with dark circles.

"It's just metadata," she told her reflection, her voice sounding thin in the quiet apartment. "It's just corporate metadata. They over-engineered the vector."

She dried her face with a hand towel and walked back to her workstation.

If the second block was binary data, it meant it was either an encrypted string, a compressed file, or a set of structural instructions.

She looked at the architecture of Chromosome 4 in her visualization tool. In human genetics, the spatial organization of DNA inside the nucleus—how it folds, loops, and twists around histone proteins—is just as important as the sequence of letters itself. A string of DNA doesn't just sit like a long ribbon; it folds into complex three-dimensional structures called chromatin loops.

If you change the base sequence of a region, even an intron, you can accidentally alter the physical shape of the local DNA. You can create new folding points, or bring distant regions of the chromosome into physical contact with each other.

Eden loaded a 3D chromatin folding predictor—an open-source deep learning model called AlphaFold-C that predicted chromosome conformation based on primary nucleotide sequences.

She loaded the natural human reference genome for Chromosome 4 into the predictor and rendered the local 3D architecture around the Huntington's gene.

The computer rendered a smooth, elegant hairpin loop. The natural non-coding intron folded back on itself gently, allowing regulatory proteins to access the neighboring promoter regions without friction.

Then she fed her own personal sequence—watermark, second anomalous block, and all—into the predictor.

The GPU fans on her workstation spun up into a high-pitched whine as the machine calculated the spatial energy landscapes and electrostatic interactions across billions of molecular configurations.

A minute later, the 3D model refreshed on her screen.

The smooth hairpin loop was gone.

In its place, the computer rendered an sharp, unnatural right-angle bend. The two anomalous base pair blocks—the text watermark and the compressed binary payload—acted like two magnetic blocks embedded in a flexible rope. They were complementary to each other.

When the DNA strand folded inside her nucleus, the first block snapped directly onto the second block, forming a rigid, artificial double-stranded stem-loop structure.

It was a genetic hairpin switch.

In molecular biology, a synthetic stem-loop is a physical mechanism. It acts like an anchor or a traffic barrier. When the cell's RNA polymerase moves along the chromosome to transcribe genes, it glides smoothly down the DNA ribbon until it encounters a hairpin loop.

If the loop is small, the polymerase knocks it open and keeps going.

If the loop is rigid, stable, and engineered with high GC-pair hydrogen bonding, the polymerase stalls.

Eden stared at the rendered 3D structure. The artificial hairpin sat precisely seventy-five base pairs upstream from the transcription start site of HTT—the gene responsible for coding the huntingtin protein, an essential protein required for the survival of neurons in the central nervous system.

If that hairpin was open, the cell functioned normally. The polymerase moved through, transcribed the gene, and the brain survived.

If that hairpin snapped shut, transcription would hit a brick wall. The cell would stop producing functional huntingtin protein.

Without huntingtin protein, striatal neurons undergo rapid, massive apoptosis. The brain disassembles itself from the inside out within weeks—mimicking the final, fatal stages of advanced Huntington's disease, but at a speed fifty times faster.

Eden sat slowly down on the edge of her desk, her knees feeling suddenly weak.

The two hundred anomalous base pairs weren't just a corporate copyright tag.

They were half of an engineered biological trigger.

By 6:00 AM, the city below was beginning to wake up. The low hum of early morning commuter traffic echoed off the damp pavement outside.

Eden hadn't slept. She had spent the last two hours analyzing the thermodynamic stability of the hairpin loop in her genome.

Her code showed that under normal physiological conditions—body temperature at 37 degrees Celsius, standard intracellular pH, normal salt concentrations—the hairpin remained in a meta-stable, unlatched state. The two anomalous blocks hovered near each other inside her cells, close enough to alter the local folding angle slightly, but not quite locked together.

The switch was currently set to OFF.

She was healthy. Her brain was fine. She had been living with this meta-stable trigger in every single cell of her body for twenty-four years without experiencing a single symptom.

"What locks it?" she muttered, staring at the thermodynamic equations scrolling across her secondary display. "What makes it snap shut?"

To lock a nucleic acid hairpin, you needed a molecular catalyst—a specific small molecule, a short single-stranded RNA sequence, or a change in local chemical environment that lowered the activation energy required for the two strands to bind together permanently.

In synthetic biology, this was known as an in vivo riboswitch. It was a standard tool in bio-engineering laboratories when scientists wanted to control gene expression in experimental mice or cell cultures. You gave the mouse a specific drug in its drinking water, the drug bound to the riboswitch, the switch changed shape, and the target gene was instantly turned off or on.

It was an elegant, precise, and completely controllable kill switch.

Except Eden wasn't a laboratory mouse. She was a human being walking through the streets of a major American city.

She opened her phone and called Dr. Marcus Hayes’s direct line at the clinic.

It rang four times before going to his personal voicemail. "This is Dr. Hayes. I’m currently away from my desk or in consultation. If this is a patient emergency, please call the main Nexus Biome diagnostic extension..."

She hung up without leaving a message.

Calling Marcus directly was a mistake. If Marcus knew about this—if he was part of the team that monitored the zero-series genetic assets—calling him on an unencrypted mobile line would immediately raise a flag in Nexus Biome’s internal network security monitoring system.

She needed to know if this switch was unique to her, or if it was baked into the core delivery vector used for all prime-edited patients.

She grabbed her coat, swept her thumb drive off the desk, and jammed it into her pocket.

She needed to talk to someone who didn't owe their mortgage to Nexus Biome.

The Institute for Evolutionary Genomics was located five miles south, on the edge of the university medical campus. It was an old, red-brick building that smelled of damp concrete and autoclaved glassware, a stark contrast to the gleaming, scented towers of Nexus Biome.

Eden walked through the double glass doors at 7:30 AM, bypassing the sleepy security guard at the front desk with a wave of her old student proxy badge.

She took the stairs down to the basement, where the low-budget bioinformatics labs were situated.

In the far corner of the lower concourse sat a room labeled Comparative Genomics & Phylogenetics Lab. Inside, surrounded by towers of discarded server racks and overflowing recycling bins of printouts, sat Dr. Aris Thorne.

Aris was thirty-two, wore faded flannel shirts regardless of the weather, and possessed the posture of someone who had spent his entire adult life leaning into a 27-inch monitor. He had been Eden’s lab partner during her master’s program before he abandoned environmental modeling to pursue open-source genomic sequencing.

"You look terrible," Aris said without looking up from his keyboard as Eden walked into the lab. He was holding a lukewarm mug of instant coffee.

"I need you to look at a read alignment," Eden said, dropping her aluminum thumb drive onto his desk.

Aris blinked, finally turning his head to look at her. "Good morning to you too, Eden. No 'how are you, Aris', no 'how's the low-pass sequencing project going'?"

"Aris, please. Load the file."

Recognizing the flat, rigid tone in her voice, Aris stopped smiling. He picked up the drive, plugged it into his local Linux terminal, and mounted the encrypted volume she pointed to.

"What am I looking at?" he asked, dragging the FASTQ file into his local alignment viewer.

"Chromosome 4. Position 3,076,604."

Aris typed the coordinates into his command line. The visual browser popped up, showing the familiar reference genome align against her dataset.

He leaned in closer to his screen, his eyes scanning the base pair tracks.

"That's a clean edit on the CAG repeat," he noted casually. "Standard Nexus prime delivery. Very low noise..." He stopped mid-sentence. His hand reached for his glasses, pushing them up his nose.

He zoomed in on the downstream intron region.

"What is that?" Aris asked, his voice dropping an octave.

"That's my blood," Eden said.

Aris scrolled through the two hundred base pairs of the ASCII watermark, then moved further down to the eight-hundred-base-pair binary payload block. He didn't speak for nearly two full minutes. His fingers flew across his keyboard, running a local structural prediction script similar to the one Eden had run hours earlier.

When the 3D model rendered on his screen, showing the tight, artificial stem-loop hairpin hanging over the HTT promoter, Aris spun his chair around to face her.

"Eden," he said, his voice barely a whisper. "Who designed this vector?"

"Nexus Biome. 2028. The Prime Edit Zero protocol."

"This isn't an off-target artifact," Aris said, pointing a finger at the screen. "Look at the GC distribution on the stem margins. It's engineered with a melting temperature of forty-two degrees Celsius. It's hyper-stable. It's a synthetic riboswitch."

"I know."

"It's targeting the HTT locus," Aris continued, his eyes wide with a mix of academic fascination and sudden terror. "If this hairpin locks, it completely silences the gene. You'd suffer acute neurodegeneration in weeks. It's... it's a kill switch, Eden."

"Is it in the public reference databases?" she asked. "Is it in any of the open-access patient sequences from Nexus?"

Aris immediately pulled up the 1000 Genomes Project archive and the public NIH Sequence Read Archive. He wrote a grep command to search millions of public human genomic datasets for the exact sequence of the ASCII watermark: PROPRIETARY: NEXUS-BIOME-PAT-2028-001.

The terminal swept through gigabytes of public data.

Matches found: 0.

"It's not in the public database," Aris said. "None of the open-source human sequences have it."

"Search for the binary payload block," Eden insisted. "Just the eight-hundred-base-pair hairpin driver."

Aris copied the sequence string and launched a broader search across public clinical trials data submitted by Nexus Biome between 2028 and 2035.

The screen flickered as the search processed.

Suddenly, a single hit popped up on the monitor.

Not from a human patient file.

It was from a patent filing archive submitted to the United States Patent and Trademark Office in November 2028, classified under Proprietary Biological Safety & IP Protection Protocols.

Aris clicked the link. The PDF document opened slowly, revealing a heavily redacted corporate patent application filed by Nexus Biome Legal Counsel.

At the top of the page, beneath the corporate logo, sat the official title of the construct:

System and Method for Conditional Genomic Expression Suppression in Transgenic Biological Assets.

Eden leaned over Aris’s shoulder, her eyes fixed on the patent abstract.

The document described a synthetic genetic construct designed to be inserted alongside high-value therapeutic edits. Its stated purpose was unambiguous: to allow the patent holder to remotely or chemically induce targeted gene silencing in the event of unauthorized distribution, genetic theft, or breach of proprietary licensing agreements.

Eden stood in the dim light of the basement lab, listening to the hum of the server fans.

She wasn't just the first human born without Huntington’s disease.

She was a patented asset with an active breach-of-contract device written into her DNA.