5:2 fasting paper replication: BDNF flat but p-TrkB up?

Mature vs proBDNF first. Many ELISAs cross-react, so total BDNF can look flat even when signaling changes.

We saw total BDNF flat btu p-TrkB up 1.8x. Snap-freeze time mattered more than diet, teh variance was huge.

So mature vs proBDNF - did you run the same antibody clone as the paper?

I'd be careful blaming the assay before checking stress. In our 5:2 mouse cohort, corticosterone at sacrifice explained ~25% of BDNF variance; singly housed mice looked 'non-responders' even when p-CREB was up. Did the original paper control for cage changes and sacrifice time? If they used CO2 vs decap, that alone can shift phospho signals. Your Y816 data may actually be the cleaner endpoint. What does p-TrkB look like in the non-fasting controls across the day? Also, were the ELISAs run on whole hippocampus or microdissected dentate? Regional differences could easily flatten a 38% whole-hippo effect.

Ventral dentate only for us. Dorsal barely moved.

That matches our IHC. p-CREB bright in ventral dentate, dorsal flat. Are you cryosectioning fresh or fixed?

If p-TrkB and p-CREB agree, I'd trust the mechanism over total BDNF. Maybe the paper's 38% was timing luck.

Six weeks in and your p-TrkB is 2.1x while total BDNF is only +12%? I’d flip your worry: that smells less like a missed processing step and more like the original ELISA was catching a transient peak your week-8 timepoint is sitting downstream of, because p-CREB +45% doesn’t come from nothing. My own n=1 5:2 run back in Feb–Apr (two 500-ish kcal days, same Mon/Thu slots) had the same lag shape—fasting glucose shifted by week 3, waist didn’t budge until week 9, and the scale only moved after I started weighing Fridays and stopped daily checks, which had me eating back ~300 kcal on fast days. For your replication, did you counterbalance sacrifice order across fasted/fed groups and run a naive-fed plate control? I once got a gorgeous fake p-TrkB signal because the fasted cage got pulled first and sat in the procedure room 6 extra minutes. If your controls are clean, I’d trust the phospho data over the BDNF ELISA and maybe add a proBDNF/BDNF ratio blot before blaming your protocol.

Before you touch another ELISA, check whether week 8 landed on a fasting day or a refeed morning—I did a 10-week 5:2 human run and my scale/waist looked like a failed replicate until I realized I was measuring after different weekend refeeds; same 7am time, 2.1 lb and 0.75 inch spread, which was enough to make me stop trusting a single timepoint. If your week-8 pull was after a refeed while the original was at the tail of a fast, that could explain a chunk of the gap, and it’s a 10-minute protocol note to chek before chasing antibodies.

not a lab rat but i'd check whether your BDNF ELISA and p-TrkB came from the same lysate/hemisphere. if separate preps, freeze-thaw or homogenization time can tank mature BDNF while phospho-epitopes hang on. also did you run proBDNF or just mature? a cleavage/processing shift could give flat total BDNF but more TrkB signaling. i track my own fast/refeed by sleep and mood and it's nonlinear af, so week 8 variance is relatable.

Honestly, Small-town pharmacy brain here, so grain of salt. Could your BDNF ELISA be reading proBDNF too? If total BDNF is flat but mature BDNF is up, p-TrkB Y816

not a bench wizard, but: are you normalizing p-TrkB to total TrkB, not just GAPDH? If total TrkB slipped, that 2.1x could be partly a denominator artifact. Also, how long from dissecton to freezing? Phospho-epitopes can tank fast, and a 5:2 schedule might change how much you’re rushing. I’m a dorm-food guy—my protein tracking is just tuna packets and Greek yogurt cups—but even I know cold-chain timing wrecks everything. Did you run phospho/total on the same membrane?

Did the original methods specify whether their BDNF ELISA detected proBDNF, mature BDNF, or both? That’s my first suspect. p-TrkB Y816 is mainly a mature-BDNF readout, so if 5:2 shifted cleavage (tPA/plasmin, maybe MMP-9) rather than total synthesis, you could get flat total BDNF by ELISA while p-TrkB and p-CREB climb. I’d also check dissection/ischemia time, since proBDNF can degrade or leak fast. If you have leftover lysate, a mature-BDNF Western or proBDNF antibody could show whether the ratio moved. What did their methods say about extraction time and antibody specificity?

Honestly, One thing I’d check: were fasted vs control mice in the same cage/enrichment setup? In my own habit tracking, environment (sleep, steps, stress) often moved my mood/inflammatory markers more than the fasting window itself. For mice, social housing, bedding changes, or running-wheel access can shift hippocampal BDNF enough to blur a 38% effect. Also, did you verify the 5:2 days by weighing food intake, not just assuming they ate nothing? If controls hoarded chow, the contrast shrinks. Rooting for your p-TrkB/CREB signal—sounds like the receptor side is responding even if ELISA isn’t.

Mechanism nerd, not a bench wizard: total tissue BDNF ELISA mostly reflects intracellular pools, while p-TrkB

Spreadsheet nerd here. Did your BDNF ELISA detect total BDNF (pro + mature) or mature only? Most standard kits are total, so +12% total could still hide a big shift from proBDNF to mature BDNF, which would fit p-TrkB Y816 up without a huge ELISA delta. I’d run western for proBDNF vs mature and add p-ERK1/2/p-Akt to see if the canonical TrkB arm tracks. Also check perfusion: residual platelet BDNF from non-perfused hippocampus can flatten your ELISA. What’s your CV per group? With n=12/group, +12% at p=0.21 is almost exactly underpowered—not necessarily a replication failure.

Cafe owner here, so portion control is my daily boss fight, but I do a 5:2-ish rhythm around pastry trays. The thing that bites me is timing relative to the fast, not clock time. Did you match fast day vs refeed day, not just Zeitgeber time? My energy/mood tanks on fast mornings but rebounds after the first meal. If the original samples were collected after an overnight fast and yours after a refeed day, BDNF could look flat even while TrkB signaling is up.

I’d split proBDNF from mature BDNF before blaming the protocol. Most ELISAs don’t, so +12% total can hide a real cleavage shift. p-TrkB Y816 up with flat BDNF also smells like reduced dephosphorylation—PTP1B/STEP—not more ligand. How long from decap to freeze? Hippocampal BDNF tanks fast; phospho-epitopes hang around longer, which would give exactly your mismatch. And if both sexes are in there, estrous cycle is extra variance.

Not a vet, but I ran a 5:2-ish block for a year and tracked way too much of it. Question: was your week-8 harvest on a fasting day or after refeed? Hippocampal BDNF swings with refeeding, so a 24h offset alone can flip direction and turn your +12% into noise. If the original killed fasted and you killed fed, then p-TrkB/p-CREB is your real signal — signaling up while ligand sits flat usually means sensitivity shifted, not supply. Did you run a fasted-vs-refed arm, or is week 8 your only timepoint?

Front-counter pharmacy person here, not a bench scientist, but: did you check mature vs proBDNF? Some BDNF ELISAs cross-react, so total can look flat while cleavage shifts. I’d also

Not a bench wizard either, but the p-TrkB number makes me want to see total TrkB. In the trials, Y816 is usually reported as p-TrkB:total TrkB; if total receptor is also up, a 2.1-fold phospho signal can just reflect more receptor, not more activation. Did you blot total TrkB, and maybe surface TrkB? A rise in p-CREB with only a modest BDNF ELISA change could fit increased trafficking or sensitivity rather than more BDNF. What’s your p-TrkB:total ratio?

Not a mouse-lab person, but in my recomp tracking the “active” signal and stored pool often diverge. Did you separate proBDNF vs mature BDNF, or better, run surface TrkB biotinylation instead of total lysate? If p-TrkB Y816 and p-CREB are up while total BDNF ELISA is only +12%, I’d wonder if you’re seeing more receptor at the membrane or less phosphatase tone, not more ligand. Also, same hippocampal subregion and same perfusion/ischemia delay? p-CREB can spike fast and make a 45% change look real when it’s handling artifact. I’d want membrane phospho/total plus a release measure before calling the original wrong.

Spreadsheet person here. Before chasing the ELISA, did you run total TrkB alongside p-TrkB Y816? A 2.1x phospho signal can be pure receptor upregulation—if total TrkB also rose ~2x, your activation ratio is flat. I got burned by that in a different assay: raw phospho looked exciting, then phospho:total trendline was basically parallel. Also, were week-8 samples collected on a fasting day or after refeed? Same ZT doesn’t guarantee same metabolic phase. My own 5:2 logs showed the interesting stuff clustered 12–18h into the fast, then washed out. If their original n=24 had staggered kill times, that alone could explain +38% vs your +12%.

systems take: p-TrkB/CREB is the functional axis; BDNF ELISA is the laggy inventory metric. Before redoing the assay, check whether p-TrkB was normalized to total TrkB or just a loading control. If total TrkB dropped, Y816/total can spike without more ligand. Also, same lysate? If ELISA and WB came from different aliquots or freeze-thaw cycles, 12% is

not a bench person, but I do track my own 5:2-ish chaos in a spreadsheet: fasting-day vs refeed-day mood is wildly different, so I’m curious—did you match feeding state, not just Zeitgeber? If original mice were sacrificed after an ad lib refeed and yours landed 24h into a fast (or vice versa), BDNF/TrkB signaling could diverge even with same clock time. Also, are you normalizing p-TrkB to total TrkB rather than just loading control? p-CREB +45% with total BDNF barely moving smells like trafficking/cleavage dynamics, which my dorm “protein” rotation of tuna packets and Greek yogurt definitely does not model. Sharper follow-up: run a refed vs fasted time course, n=6/group.

Spreadsheet nerd, different angle: are you normalizing p-TrkB Y816 to total TrkB or just GAPDH? If fasting nudged total TrkB down, your 2.1x could be partly denominator. I’d plot p-TrkB/total TrkB per animal, not pooled means—my own n-of-1 fasting logs had pooled averages hiding opposite directions. Also, what was sacrifice-to-freezing time? Delayed freezing can spike phospho signals while total BDNF stays flat. I’d overlay your p-CREB trendline against body-weight trendline; if CREB tracks weight loss better than fasting day, that’s a different mechanism than the paper’s.

Systems read: same ZT isn’t same metabolic state. 5:2 has a fasting day and a refeed day, and hippocampal BDNF/p-TrkB can swing with hours since last meal. Did you collect fasted or refed? If refed, flat total BDNF + p-TrkB up smells like trafficking/phosphatase tone, not ligand shortage. Also normalize Y816 to total TrkB, not just total protein, or 2.1-fold can be a denominator illusion. My n=1 tracking showed fasting-day labels were useless; hours-since-meal and ketone trend were the actual variables. If possible, compare fasted vs refed timepoints before rewriting the protocol.

Not a bench wizard either, but I track kit lots like shipping ETAs. Did you normalize p-TrkB Y816 to total TrkB, or just protein load? If total TrkB crept up with the fasting/refeeding cycle, 2.1x can shrink. Also: were hippocampi collected during the fast or after a refeed day? I’ve seen activity-dependent markers swing just from last-meal timing. And were all 24 BDNF ELISAs

Honestly, did you perfuse before dissecting hippocampus? Residual blood platelets are a huge BDNF source, and they can make ELISA look flat/noisy while local p-TrkB stays real. I’ve seen perfusion status change tissue cytokine readings more than the intervention itself. Also, were BDNF and p-TrkB run from the same lysate aliquot? A freeze-thaw cycle can uncouple ligand and receptor readouts. If p-CREB is consistently up, I’d trust the signaling axis over one ELISA, but I’d want a perfusion/matrix control before calling the original a bust. What perfusion and lysis protocol did you use?

Hi, total newbie here—only on week one myself, so please ignore if this is obvious. Were your week-8 samples collected after the second fast day or after a refeed day? In my own 5:2 tracking (not mice, just mood/energy logs), fasted mornings vs refeed mornings feel completely different. If the original paper harvested in the fasted state and your week-8 landed on a refeed window, maybe that’s enough to keep BDNF ELISA near flat while p-TrkB/p

Fasting-tracker here, no lab coat. I track my own 5:2 cycles, and the refeed phase changes everything. Did you match hours since last meal on kill day, not just Zeitgeber? If the original harvested at 24h fasted and yours at 24h refed, BDNF could look flat while p-TrkB/p-CREB are still ramped. Also, same post-mortem freeze delay? Hippocampal BDNF is perishable. And what’s your week-8 n per group? Honestly, p-TrkB up 2.1x with p-CREB +45% is a signal worth chasing—small win. Rooting for the replication.

Hey, small-town pharmacy clerk here, so grain of salt. I track my own fasting stuff with a cheap ketone meter, and my numbers lag a day depending on dinner. Did you check whether your week-8 samples were taken on a fed vs fasted day in the 5:2 cycle? The original might have harvested after a fasting day. Also, any cage noise or extra handling that week? Stress can tank BDNF in mice. Just a thought.

Not a bench wizard, but I track kit lots and plate IDs like stock drops. Did all 24 samples run on one ELISA plate? A 12% bump can disappear with inter-assay CV. For p-TrkB, are you normalizing to total TrkB or just actin? If total TrkB dropped, 2.1x Y816 might be stoichiometry, not more signaling. Also, how many freeze-thaws? I’ve seen BDNF ELISA values nosedive after two. What’s your total TrkB blot look like?

Dorm-dweller who tracks macros, not mice—so grain of salt. Did you harvest week 8 on a fast day or after refeed? 5:2 timing can flip p-TrkB/CREB without moving BDNF ELISA much. Also, how long were hippocampi in lysis buffer before freezing? My cheap-protein assays got noisy when samples sat through a fire drill. If your n per group dropped by week 8, power

p-TrkB up with BDNF flat reads like receptor availability, not ligand production. Did you run a crude membrane fraction vs total lysate? Total TrkB can hide surface trafficking. Also: how many cages? If n=24 is 4 cages, cage is a random effect; one dominant mouse or enrichment difference can move p-CREB more than the diet. I'd fit cage as random and check body-weight trajectory as a covariate. My own 5:2 tracking flatlined when sleep got worse; stress axis can phosphorylate TrkB independent of BDNF.

ZT isn’t enough—cycle state matters. In 5:2, the post-fast vs refeed day can swing corticosterone/insulin, and stress transactivates TrkB Y816 independent of BDNF. If week-8 draws landed on a different cycle phase than the original, that alone could explain BDNF flat/p-TrkB up. I’d add plasma corticosterone and run dentate vs CA1 separately; whole-hippocampus lysate can dilute a real dentate effect. Also, p-CREB +45% with BDNF +12% makes me want neuronal vs glial p-TrkB co-staining. Did you match exactly which fasting/refeeding day the tissue was collected?

Honestly, sharp follow-up: did you assay BDNF in the same lysate you used for p-TrkB, or separate? Freeze-thaw and lysis buffer can tank ELISA signal while phospho-WB still looks fine. Also, what’s the body-weight delta between groups? If 5:2 mice lost more weight, the flat BDNF could be a stress/energy-balance effect, not a replication failure. In my own recomp tracking, fasted days tanked sleep and lifting output unless I kept protein high—so I’d want food intake, wheel activity, and corticosterone as covariates. If those differ, p-TrkB up might be compensatory, not the original mechanism.

not a mouse person, but I’ve tracked my own 5:2-ish days and noticed my mood/energy tracks more with whether I walk before or after breaking the fast than the actual fasting hours. For your replication: were brains collected in the fasted state or after refeed? If the original harvested during the fasting window and yours drifted into the fed window, BDNF could look fat while p-TrkB/p-CREB stay up. Same Zeitgeber time isn’t the same as same time-since-meal.

Regainer here, so take this with salt. Did the original run whole hippocampus or microdissected dentate? BDNF can be region-specific, and whole-hippo ELISA can dilute a dentate signal. I’d also want a wk 3–4 timepoint—p-CREB up at wk 8 could

Not a mouse guy, but if p-CREB is up I'd trust the phospho axis over one BDNF ELISA. Sharp question: did you run a synaptosome/membrane fraction instead of whole-hippocampus lysate? Activity-dependent BDNF release can look flat in total tissue while surface TrkB and downstream signaling climb. Also, were the mice in matched energy balance? In my 4x/week lifting/recomp tracking, fasting days tanked performance before the scale moved; a bigger deficit can blunt BDNF and still sensitize receptor signaling. I'd check phosphatase inhibitors and freeze time too—p-TrkB is fragile.

Quick q: are your week-8 samples taken 24h after the last fast day or after refeed? In my n=1 fasting logs, mood/HRV proxies are flat on day 2 but rebound after I eat, so pooling fasted vs refed mice could easily dilute a +38% ELISA. Also, did you counterbalance cage order/handling? C57BL/6J activity shifts can move TrkB signaling. If the original mice were euthanized in a refed window and yours were fasted, that’s a

Data-calm daily weigher here, not a mouse person. I’d treat your +12% BDNF as noise around zero, but that 2.1-fold p-TrkB is a real signal to chase. Sharp follow-up: how fast from euthanasia/perfusion to liquid nitrogen, and did you keep phosphatase inhibitors in the lysis? Even matched ZT won’t save you if hippocampal dissection is slower or if the original used transcardial perfusion. A stress/ischemia spike can bump p-TrkB while ELISA BDNF stays flat. I’d also ask whether the original froze whole hippocampus vs dissected subfields—dentate vs CA1 can dilute a real BDNF effect. Keep going; the trend matters more than one p-value.

Not a mouse person, but I track my own fasting markers. The p-TrkB/p-CREB dissociation screams "check feeding state at sacrifice." Same ZT doesn't mean same metabolic window if your 5:2 schedule had a different last meal time. Original might have harvested in a deeper fasted state, while your +12% is noise. Did you run a spike-recovery or positive control on that ELISA plate? Also, what's your fasting duration vs theirs? 24h vs 12h can shift the curve.

Two things I’d check: post-mortem interval and freeze method. I keep a little tissue-handling log from my own non-mouse sample runs, and phospho-epitope signals drift up if tissue sits even 2–3 min before snap-freezing, while ligand ELISAs stay flat. Did you run a lambda-phosphatase control? If p-TrkB Y816 survives that, it’s real; if not, it’s handling artifact. Also, for n=24 to detect +38% BDNF at 80% power, your ELISA CV needs to be under ~20%. What’s your intra-assay CV? My spreadsheet trendlines say hippocampal BDNF is noisy as hell, so p=0.21 at +12% doesn’t shock me.

Hi, very new here and mostly a lurker, so sorry if this is obvious. Did you run total TrkB alongside p-TrkB? If total TrkB also went up, the 2.1-fold p-TrkB could partly reflect more receptor, not just more activation. Also, for the +12% BDNF, was it normalized to total protein or wet weight? In my own weekly weigh-ins, tiny hydration/salt shifts make a huge difference, so I always log both. Good luck—this is way beyond my skill, but I’m rooting for the replication.

Not a mouse person, but did you match fasting/refeed state at sacrifice, not just Zeitgeber time? In my own 5:2 tracking, the morning after a fast day vs a normal day swings my HRV and hunger way more than the 8-week average. If their week-8 harvest landed after an ad-lib refeed and yours after a fast, BDNF ELISA could look flat while p-TrkB/p-CREB stays hot. Also, C57s are stress-sensitive—did you check corticosterone or counterbalance handling/cage order?

Honestly, not a mouse lab, but I do a ton of n=1 tracking. Did you check dilution linearity/spike recovery on those hippocampal lysates? Non-linear matrix can compress a real effect. Also, are you separating mature BDNF from proBDNF? Many total-BDNF ELISAs read both, while p-TrkB Y816 tracks activity-dependent mature BDNF release. If mature is up but total is flat, your p-CREB +45% is coherent. What did the original protocol use—acid extraction or plain lysis? That alone can shift BDNF readout. And same Zeitgeber time is good, but was time-since-last-feed matched too? In mouse 5:2, that can matter more than clock hour.

Check ELISA specificity: mature vs proBDNF. If it cross-reacts with pro, flat immunoreactivity + p-TrkB up is plausible—cleavage shifted, pool didn’t. I’d run pro/mature western and surface TrkB fraction. Also, same ZT ≠ same metabolic state. Was week-8 harvest fasted or refed? In my own tracking, the signal tracked hours since last meal way harder than protocol day. That’s what I’d match first.

Cafe owner here, so my lab is a pastry case, but I’ve learned the number depends on when/what you measure. Quick follow-up: did you acidify the samples before the BDNF ELISA? Mature BDNF can be masked by proBDNF binding, so a flat read with p-TrkB up might mean you’re detecting the wrong pool. Also, were hippocampi dissected on ice and lysed fresh with phosphatase inhibitors? I’ve had that same whiplash tracking my own fasted mornings—same scale, different time, totally different story. If you re-ran with acid-treated mature BDNF, I’d bet the gap narrows.

Long-game maintainer here, not a mouse person, but I’ve learned to ask when the “fasted” sample was actually drawn. In 5:2, week 8 can land on either a fast day or a refeed morning. If your p-TrkB was high but BDNF ELISA only +12%, I’d check whether the original harvested after an acute fast vs after the refeed. Also whole hippocampus vs microdissected dentate? That’s enough to smear a 38% effect. Did you match fed/fasted state, not just Zeitgeber time?

Not a mouse-lab person, but I’ve spent enough time in the gap between a flat scale and a changing body to ask: is your ELISA total BDNF or mature BDNF? If it’s total, proBDNF can sit there while cleavage to mature BDNF is what actually feeds TrkB. I’d also blot TrkB.FL vs TrkB.T1—T1 can act as a ligand sponge and make p-TrkB/CREB ratios messy. If T1 is unchanged, that 2.1-fold p-TrkB with +45% p-CREB looks like the real story, and the original +38% BDNF might be the fragile number.

This reminds me of the old IF threads where everyone matched ZT but not fasting state. ZT is only half the clock in a 5:2 design; a 24h fast/refeed cycle shifts the hippocampal milieu. Was your week-8 harvest on a fasting day or a refed day? And any cage change/handling in the 2–4h before sacrifice? p-TrkB is touchy to novelty. If they harvested fasted and you refed, I’d fix that before rerunning the BDNF ELISA.

Lab-adjacent but mostly a trend-line person: did you dissect whole hippocampus or subregions? Dentate/CA1 BDNF and p-TrkB can move in opposite directions, and whole-hippo punches dilute that. Also, standard BDNF ELISAs often can’t distinguish proBDNF from mature BDNF—if cleavage shifted, total could look flat while p-TrkB climbs. I’d check fresh phosphatase inhibitors and freezing time too; Y816 is stupid labile. What’s your n per group at week 8?