Depression threshold levers — the low-coordination operating-point restoration, decomposed into three DNA-grounded levers (L3-dominant)

The previous chapter's operating-point restoration is decomposed into a DNA-grounded three-lever target map: modulate inward glutamatergic current (L1), increase outward potassium current (L2), or remove the upstream HPA drive and restore the deficient monoamine and neurotrophic drives (L3, dominant here). Eighteen depression genes are placed by their promoter switch stiffness; targets are ranked, never drugs or doses; not medical advice.

The depression chapter read major depression as the chronification of a low-coordination operating point: a sustained, HPA-driven withdrawal pulls the global order parameter R below the band health lives in, and with plasticity that excursion writes a retained structural trace. That chapter proved the pole and the corrective sign — depression sits at the opposite pole from epilepsy, so the therapeutic push is whatever restores the deficient drive (or removes the chronic-stress drive that lowers the operating point) — but it left that push abstract: it said restore coordination without saying which genes or drives actually realise the restoration. This chapter fills exactly that gap, with the same piece of inherited technology the bipolar and epilepsy levers chapters used: the threshold-shift intervention logic from the analgesic reproducibility package (Zenodo 10.5281/zenodo.20733420). Its three abstract levers — L1 change the inward (excitatory) current, L2 change the outward (potassium / inhibitory) current, L3 change the up-stream drive — carry over unchanged, but depression is where the frame's generality is genuinely tested, because it is the first L3-dominant case: of eighteen depression genes, twelve sit on L3, and only three each on the channel levers. The L3 set splits into three sub-axes — the HPA-removal trio (NR3C1, CRHR1, FKBP5), the monoamine-restoration set (SLC6A4, SLC6A2, MAOA, TPH2, HTR1A, HTR2A, COMT), and the neurotrophic-restoration pair (BDNF, NTRK2) — while L1 carries the glutamatergic / calcium set (GRIN2A, GRIN2B, CACNA1C) and L2 the outward set (KCNQ2, KCNQ3, GABRA1). Each gene is placed by reading its own promoter switch stiffness — γ = −mean nearest-neighbour stacking free energy (SantaLucia 1998) over its promoter window, turned into |h_sp| = spinodal(γ) with the frozen engine read-only, and seven of the reads (NR3C1, CRHR1, GRIN2A, CACNA1C, KCNQ2, KCNQ3 from the bipolar cache, GABRA1 from the epilepsy cache) are carried over verbatim. Two honest caveats are recorded, not hidden: the L1 glutamatergic direction is an NMDA antagonist (the ketamine route) acting through downstream neurotrophic signalling, not an excitation reduction, so its naive L1 sign is flagged sign-subtle; and the HTR2A direction is non-monotone (an agonist psychedelic route and an antagonist route both appear). Three fail-closed disciplines ride along: every L3 link stays graded [O] cited biology, a forbidden-claim scanner rejects any dose / efficacy / safety / synthesis statement (including mood-lift and remission language), and a burden-weighted ranking orders targets, never drugs or doses. The firewall is absolute: the promoter |h_sp| is a gene's own switch stiffness, never the §27 network operating-point on R, a receptor occupancy, a synaptic monoamine level, a potency, a dose, or a clinical effect. efficacy = 0; not medical advice; the hard problem stays open.

What §27 left abstract

The depression chapter placed the disorder at the low-coordination operating point of the atlas, built on top of the E0 plasticity layer: health is a population of micro-eddies running in parallel at a moderate global coherence, and major depression is a sustained downward excursion of the order parameter R — an HPA-driven withdrawal that holds the system below its healthy band — which plasticity then consolidates into a retained structural trace (the chronification). That chapter proved two things and stopped: the pole (depression is the under-coordination side, the mirror of the over-synchronisation disorders) and the corrective sign (the therapeutic push is whatever restores the deficient drive or removes the chronic-stress drive that lowers the operating point). That sign is forced and honest, but it describes the intervention as a single abstract operator: restore coordination. It does not say through what physical drive the restoration is realised — which axis, which receptor, which gene. A mechanism atlas should be able to say more than something restores the operating point; it should enumerate the ways an operating point can be lifted and attach real molecular targets to each. The bipolar and epilepsy chapters showed that this enumeration is exactly what the inherited threshold-shift technology supplies — and depression is where that technology is most tested, because here the answer is not a channel at all.

The inherited technology, applied a third time — and the first L3-dominant case

The handle is not invented here, and it is not adapted here — it is the same logic the bipolar and epilepsy chapters inherited from the analgesic reproducibility package (Zenodo 10.5281/zenodo.20733420), now turned on a third disorder. Its premise is general: an operating point is set by a balance of currents and the drives that bias them, so there are exactly three levers on it. L1 — change the inward, excitatory current. L2 — change the outward, repolarising (or inhibitory) current. L3 — change the up-stream drive that sets where the operating point sits in the first place. The frame applies unchanged because the depression chapter, the bipolar chapter, the epilepsy chapter, and the analgesic package all share the same R19 substrate — the engine's supercritical pitchfork ṣ = g·s − s³ + h, whose spinodal fold IS the switching barrier.

Depression makes two things vivid that the channel-led disorders did not. The first is a sign flip at the disorder level, and the model states it rather than smoothing it away: epilepsy and bipolar mania sit above health (the lever sign is reduce the excess drive), but depression sits below health, so the corrective sign is the mirror — restore the deficient drive, or remove the chronic withdrawal drive that is holding the operating point down. The three abstract levers carry over intact; what flips is their direction. The second is a shift of distribution. Where bipolar leaned on L1 (its replicated GWAS loci are calcium channels) and epilepsy leaned on L2 (the KCNQ2/KCNQ3 M-current), depression is the first case that is L3-dominant: of the eighteen genes, twelve sit on L3 — the up-stream HPA, monoamine, and neurotrophic drives — and only three each on the channel levers. This is the frame's generality genuinely exercised: the same three levers re-distribute onto up-stream biology when that is where the disorder lives, instead of forcing depression into a channel mould it does not fit. The pipeline is reused at the level of code, not analogy: seven of the reads — NR3C1, CRHR1, GRIN2A, CACNA1C, KCNQ2, KCNQ3 from the bipolar cache and GABRA1 from the epilepsy cache — are carried over verbatim, because γ is a strand-symmetric property of the sequence and does not change between problems. Nothing about the engine is touched; the module re-emerges the frozen tree read-only and confirms it byte-unchanged, and registers as the tenth atlas citizen (DEP-T1b-L).

L3 (dominant) — remove the upstream drive and restore the deficient drives

The dominant lever in depression does not touch a channel; it changes the up-stream drive that sets the operating point. Twelve of the eighteen genes sit here, and they split cleanly into three sub-axes that map onto the three great families of antidepressant mechanism. The first sub-axis is HPA removal — turning down the chronic stress drive that holds R below health. Three genes anchor it: NR3C1 (the glucocorticoid receptor; the mechanism direction is to restore its negative feedback, so the stress hyperdrive that lowers the operating point is removed), CRHR1 (the CRH receptor that initiates the axis), and FKBP5 (the co-chaperone that tunes glucocorticoid-receptor sensitivity, one of the most-replicated stress-interaction loci in psychiatric genetics). The second sub-axis is monoamine restoration — lifting the deficient serotonergic and noradrenergic drive, which is the classical antidepressant target. Seven genes carry it: the transporters SLC6A4 (the serotonin transporter, the SSRI site) and SLC6A2 (the noradrenaline transporter, the SNRI site); the catabolic enzymes MAOA (monoamine oxidase A, the MAOI site) and COMT (catechol-O-methyltransferase); the synthetic enzyme TPH2 (neuronal tryptophan hydroxylase, the rate-limiter of serotonin synthesis); and the receptors HTR1A and HTR2A. The third sub-axis is neurotrophic restoration — rebuilding the plasticity drive that lets a chronified operating point un-write. Two genes anchor it: BDNF (brain-derived neurotrophic factor) and its receptor NTRK2 (TrkB). This last pair matters out of proportion to its size, because it is the convergence point: the rapid-acting glutamatergic route and the slow monoamine route both appear to lift the operating point through downstream BDNF / TrkB signalling, which is why neurotrophic restoration tops the priority ranking below.

This is the lever that demands the most discipline, and it gets it. Every L3 link is held at grade [O] — open, cited biology, never derived from the substrate. The model does not claim to compute that the HPA axis or BDNF sets the network operating point; it records, as cited upstream biology, that these are the drives whose restoration (or, for the stress axis, removal) would lift it. A fail-closed L3-honesty gate enforces exactly this: it checks that the declared twelve-gene L3 set is present, that each member is graded [O] and carries no current-channel mislabelling, that the firewall sentence separating the promoter |h_sp| from the §27 network operating-point is present, and — the check unique to this chapter — that L3 is genuinely dominant (more L3 members than L1 or L2). It FAILS the build if any of these slips. And the HTR2A entry carries the chapter's second honest caveat openly: its direction is non-monotone — both an agonist route (the serotonergic-psychedelic direction) and an antagonist route (the direction shared with several sedating antidepressants) appear in the literature, so a single signed lever cannot capture it, and the model records the ambiguity rather than picking a side. A direction, never a dose — efficacy = 0.

L1 — the inward glutamatergic / calcium levers (and the ketamine subtlety)

The first channel lever is minor in depression — three genes — but it carries the chapter's first and most important honest caveat. GRIN2A and GRIN2B are NMDA-receptor subunits (the GluN2A and GluN2B glutamatergic, calcium-permeable channels), and CACNA1C is the L-type calcium channel Ca_V1.2 (the same top-replicated cross-disorder locus that anchored the bipolar L1). The naive reading of L1 would be reduce inward excitatory current — but in depression the most important glutamatergic agent, ketamine (and esketamine), is an NMDA antagonist, and its rapid antidepressant direction is not a simple excitation reduction. The current understanding is that NMDA antagonism on inhibitory interneurons produces a downstream disinhibition and a surge of BDNF / TrkB signalling — i.e. the L1 agent acts by lifting the L3 neurotrophic drive, not by quieting excitation at the operating point. The model flags this explicitly as a sign-subtle entry: the lever placement (the gene is on the inward-current lever) is structural and trait-blind, but the direction of the clinically relevant agent runs through a downstream route and is gene-, dose-, and mechanism-specific, and stays graded [O]. This is exactly the case the firewall exists for: a promoter read places a gene on a lever; it says nothing about whether an agonist or an antagonist is the therapeutic direction. A direction, never a dose.

L2 — the outward potassium / GABA-A levers

The second channel lever is also minor in depression — three genes — and it is the direct mirror of the inward levers: strengthen the outward, restoring side. KCNQ2 and KCNQ3 are the neuronal M-current pair (K_V7.2 / K_V7.3) — the same channels that were the dominant lever in epilepsy — and here a K_V7 opener has been studied (exploratory) for the reward / anhedonia dimension of depression, so the gene is carried as a minor lever with its promoter read reused verbatim from the bipolar / epilepsy caches. GABRA1 (the GABA‑A α1 subunit, its read reused from the epilepsy cache) rides L2-adjacent: it carries no potassium current, but it raises the same restoring side through inhibitory chloride conductance — the same L2 logic realised through a different ion. The contrast with epilepsy is the whole point of the distribution finding: the identical KCNQ2/KCNQ3 channels that dominate the epilepsy map sit at the periphery of the depression map, because depression's centre of gravity is up-stream, not at the firing threshold. Every claim on this lever is a mechanism direction only; no opener, blocker, dose, or patient is named.

The DNA grounding: a promoter's own switch stiffness

What places each of the eighteen genes is not a list but a read. For every gene, the module takes its promoter window (transcription start −2000 to +500 bases, Homo sapiens) and computes γ = −mean of the nearest-neighbour base-stacking free energies along that window (the SantaLucia 1998 nearest-neighbour thermodynamics), then turns that γ into the promoter's switch stiffness through the frozen engine's own functions: |h_sp| = spinodal(γ) = 2(γ/3)^1.5 and barrier = γ²/4. The reads span a real range — the stiffest promoter in the set is KCNQ2 at γ ≈ 1.58 (|h_sp| ≈ 0.76), closely followed by the HPA-axis gene CRHR1 (γ ≈ 1.56, |h_sp| ≈ 0.75) and the serotonin transporter SLC6A4 (γ ≈ 1.52, |h_sp| ≈ 0.72); the softest is GABRA1 at γ ≈ 1.25 (|h_sp| ≈ 0.54), with the glucocorticoid receptor NR3C1 also near the soft end (γ ≈ 1.26, |h_sp| ≈ 0.55), and the neurotrophic gene BDNF reading mid-range (γ ≈ 1.43, |h_sp| ≈ 0.66). These are read on the same R19 substrate, with the same engine, that the bipolar, epilepsy, and analgesic packages used, which is the whole point of the inheritance: one substrate, one pipeline, now four problems. The γ read is a property of the gene's promoter sequence, blind to whether the gene is on or off and to gain / loss / expression level, and that is all it is.

Ranking targets, and the firewall that keeps gamma honest

The last component prioritises, and it prioritises targets, never drugs or doses. A burden-weighted score combines three declared, cited weights — clinical burden (0.40), unmet need (0.35), and genetic-evidence / druggability (0.25) — on cited 1–5 tiers, and ranks the genes by that score alone. The top actionable target is BDNF — the neurotrophic convergence point where the rapid-acting and monoamine routes meet, maximal on every tier — followed by the HPA gene FKBP5, the glutamatergic GRIN2B, and the glucocorticoid receptor NR3C1; a substantive consequence of the unmet-need weight is that it lifts these up-stream HPA / neurotrophic / glutamate targets above the high-burden but well-served monoamine transporters such as SLC6A4. The three channel genes KCNQ2, KCNQ3, and CACNA1C are flagged not actionable by the generic lever sign in depression, carrying their exploratory / indirect status right in the ranking. Crucially, the γ read is carried alongside each target as structural context but is never folded into the score — and the result is a clean demonstration of the firewall: the priority ranking and the γ / |h_sp| ranking are decoupled. The stiffest promoter read in the whole set, KCNQ2 (|h_sp| ≈ 0.76), is a minor, exploratory L2 lever and sits near the bottom of priority; the top-priority target BDNF has only a mid-range read (|h_sp| ≈ 0.66). If promoter stiffness drove the ranking, neither could sit where it does. That decoupling is the firewall made visible, and it must be stated once more in full: the promoter |h_sp| is a gene's own switch stiffness, and it is never equated with the §27 network operating-point on R, nor with a receptor occupancy, a synaptic monoamine level, a compound's potency, a dose, an in-vivo selectivity, or any clinical effect. A fail-closed forbidden-claim scanner guards the whole package: it scans the written results for any dose, efficacy-as-fact, safety-as-fact, or synthesis statement — including lifts mood, remission, relapse-prevention, and antidepressant-efficacy language — carries a negation guard, and includes a planted self-test that must fire on its own bait, failing the build if it ever does not. This module reproduces bit-for-bit with the engine byte-unchanged.

Everything here is an in-silico reading of promoter sequence and a frame for organising targets, not a clinical measure, a diagnosis, or a prescription. The model asserts mechanism directions and target placements — an operating point can be lifted three ways; these eighteen genes populate the three levers; depression loads them up-stream (L3-dominant); these targets carry the highest genetic and unmet-need burden — and nothing about which agent acts on any lever, at what dose, in whom, whether any real drug lifts anyone's mood, or that anyone should change a treatment. The agents named as directions (the SSRIs / SNRIs / MAOIs on the monoamine sub-axis, the glucocorticoid and CRH routes on the HPA sub-axis, the ketamine route flagged sign-subtle on L1, the M-current opener on L2) are illustrations of a sign, never a recommendation — and the two recorded caveats (the downstream-BDNF route of the L1 NMDA antagonist, the non-monotone HTR2A direction) are there precisely because a generic lever sign is not a clinical direction. Real depression is heterogeneous — melancholic, atypical, psychotic, peripartum, seasonal, and bipolar depression; monoaminergic, HPA, inflammatory, circadian, and psychosocial contributors; roughly a third treatment-resistant — and that heterogeneity is locked. A promoter read and a lever assignment are mechanism boundaries, not a claim about the felt quality of depression or its lifting (Axis-A firewall — consciousness_claim = 0, the hard problem stays open). This is not medical advice, not a diagnosis, not a treatment protocol, and not a cure. medium_efficacy_tested = 0; targets ranked, never drugs or doses.