Peptides and the HPG Axis

Published August 2026 · Patriot Labs Research Team · ~11 min read

What the HPG axis is

The axis has three tiers, and each one talks only to the tier below it.

Tier one — the hypothalamus. A small, scattered population of neurons produces gonadotropin-releasing hormone, GnRH, a decapeptide. These neurons do not release it into the general bloodstream. They release it into the hypophysial portal circulation, a short private vascular link running from the median eminence down to the anterior pituitary. That anatomical detail matters: it means a very small quantity of hormone can produce a high local concentration at its target without being diluted into the whole body first.

Release is episodic rather than continuous. The Endotext review of GnRH physiology describes the pulsatile mode as “episodic release of GnRH, with distinct pulses of GnRH secretion into the portal circulation with undetectable GnRH concentrations between pulses.” Between pulses the signal is effectively zero. That is not an artefact of measurement — it is the design.

Tier two — the anterior pituitary. GnRH arrives at gonadotrope cells, which express the GnRH receptor. In response they release two glycoprotein hormones into the systemic circulation: luteinising hormone (LH) and follicle-stimulating hormone (FSH). These are the gonadotropins, and unlike GnRH they travel everywhere.

Tier three — the gonads. LH and FSH act on distinct cell populations in the gonads, and their combined effect drives gametogenesis and the synthesis of gonadal steroids. So far this is a straightforward relay. What makes it an axis rather than a chain is what happens next.

Closing the loop: negative feedback

The steroids produced at tier three circulate back to tiers one and two and suppress further signalling. More gonadal output means less GnRH drive and less gonadotropin release; less gonadal output means the brake comes off. This is a classic negative feedback loop, and it is what keeps the system at a set point rather than running away.

There is a wrinkle that took decades to resolve. GnRH neurons themselves are largely blind to the feedback signal — as the Endotext review puts it, “GnRH neurons do not express estrogen receptor alpha.” If the cells that release GnRH cannot detect the steroid that is supposed to inhibit them, something else must be relaying the message. That relay is a separate neuronal population, and identifying it is the reason kisspeptin appears later in this guide.

One consequence is worth holding onto, because it explains where HMG came from. When gonadal steroid output falls away permanently, the brake is released and circulating gonadotropin concentrations rise and stay high — the loop shouting louder at a tier that is no longer answering.

Why pulsatility matters

If you were designing this system from scratch, you would probably assume more signal means more output. The axis does not work that way, and the experiment that established it is one of the cleaner pieces of endocrine physiology on record.

In 1978, Belchetz, Plant, Nakai, Keogh and Knobil published in Science the results of work in rhesus monkeys in which hypothalamic lesions had abolished the animals’ own GnRH output, collapsing gonadotropin secretion. The researchers then replaced GnRH externally in two different patterns. Intermittent replacement, timed to approximate the natural rhythm, re-established gonadotropin secretion. Constant delivery of the identical molecule did not. And when animals already restored on the intermittent pattern were switched to constant delivery, the response desensitised and down-regulated — the system that had just been working shut itself down.

The pattern of arrival, not the cumulative amount delivered, determined the outcome.

The mechanistic reading is receptor-level. Sustained occupancy of the GnRH receptor drives desensitisation and loss of responsive receptor from the gonadotrope surface; a gap between stimuli allows recovery. Take the gap away and the same ligand that sustains the axis becomes the thing that silences it. That is why one molecule can be described, without contradiction, as both a stimulator and a suppressor of the same pathway — the two descriptions refer to two delivery patterns.

The system does more than tolerate the rhythm; it reads information out of it. The Endotext review notes that rapid GnRH pulse rates preferentially increase transcription of the alpha and LH-beta subunit genes, whereas slow pulse frequency favours FSH-beta transcription. The pituitary is frequency-decoding: the interval between pulses biases which of the two gonadotropins is produced. Frequency is not noise around a set point. Frequency is the message.

Anyone reasoning about GnRH-receptor-directed molecules in a model system has to account for this. Steady-state exposure and episodic exposure of a receptor system are not two points on one curve; they are different experiments.

Kisspeptin: the upstream gatekeeper

Kisspeptin sits above GnRH. It is the answer to the question the feedback wrinkle raised, and it is also the answer to a separate question — what actually starts the axis at puberty.

The gene is KISS1, and its history is odd. It was first identified as a metastasis suppressor, which is why UniProt still records the human protein as “Metastasis-suppressor KiSS-1” and why the longest processed fragment is called metastin. The reproductive role came later.

The human precursor is 138 amino acids, proteolytically processed into a family of peptides sharing a common C-terminal region: kisspeptin-54 (metastin, residues 68–121), kisspeptin-14 (108–121), kisspeptin-13 (109–121) and kisspeptin-10 (112–121). Kisspeptin-10 is the shortest of the set and, per the UniProt record, the most studied variant. All converge on one receptor: KISS1R, a 398-residue seven-transmembrane class A GPCR whose older name is GPR54 — it was an orphan receptor before its ligand was known. It couples through G_q, activating phospholipase C and raising intracellular calcium, and also engages beta-arrestin signalling leading to MAPK/ERK phosphorylation. The hypothalamic KISS1/KISS1R system, in UniProt’s wording, “modulates the secretion of gonadotropin-releasing hormone.”

Why this is unusually solid. Most peptide stories rest on a handful of pharmacology experiments. This one rests on human genetics running in both directions. Loss-of-function mutations in KISS1R cause hypogonadotropic hypogonadism type 8, and mutations in KISS1 itself cause hypogonadotropic hypogonadism type 13 — in both cases, absent or incomplete sexual maturation with low gonadotropin concentrations. Break the receptor and the axis never starts. In the other direction, KISS1R variants are associated with central precocious puberty. That is a far stronger form of evidence than most of the compounds discussed elsewhere on this site can claim, and it is worth saying so plainly — several of the peptides we write about rest on thin, single-group literatures. This one does not.

The mechanism is also attached to specific cells. The Endotext review describes the kisspeptin–neurokinin B–dynorphin (KNDy) network as “the missing link” in steroid feedback, with human infundibular kisspeptin neurons co-expressing neurokinin B and dynorphin — one excitatory and one inhibitory co-transmitter, a plausible arrangement for generating a rhythm. The same review calls kisspeptin “the most potent GnRH secretagogue currently known.”

The kisspeptin-10 sold by Patriot Labs is that 10-residue C-terminal fragment, supplied as a lyophilised synthetic peptide in a 5 mg vial — a defined sequence that behaves like the other peptides in the catalogue. See what peptides actually are for the general chemistry.

HCG: a placental glycoprotein built from two chains

Human chorionic gonadotropin is where this guide stops being about peptides in the strict sense. HCG is a glycoprotein hormone of placental origin, described by StatPearls as “a glycoprotein composed of two subunits — the alpha and beta subunits.” Those chains are separate gene products, non-covalently associated, and neither is active alone.

The shared alpha chain. The CGA gene encodes a 116-residue protein that UniProt describes as the “shared alpha chain of the active heterodimeric glycoprotein hormones thyrotropin/thyroid stimulating hormone/TSH, lutropin/luteinizing hormone/LH, follitropin/follicle stimulating hormone/FSH and choriogonadotropin/CG.” One gene, four hormones. TSH, LH, FSH and HCG are all built on the same alpha chain.

The beta chain does the discriminating. UniProt states the position directly: the four active hormones are heterodimers of the common alpha chain “and a unique beta chain which confers their biological specificity.” For HCG that partner is the product of CGB3, a 165-residue chain. Because HCG’s beta chain is the closest relative of LH’s beta chain within this family, the two hormones converge on the same target — the receptor is named, for exactly that reason, the LH/choriogonadotropin receptor.

Why “glycoprotein” is not a pedantic distinction. The carbohydrate is not decoration. The UniProt record for the HCG beta chain lists two N-linked glycosylation sites on asparagine residues 33 and 50, and four O-linked sites on serines 141, 147, 152 and 158. Those O-linked sites sit in a disordered, proline-rich C-terminal extension spanning residues 131–165, which the record notes may enhance the hormone’s stability and bioactivity — a feature HCG has and LH does not.

Practically, this puts HCG in a different manufacturing category from every synthetic peptide we stock. A decapeptide like kisspeptin-10 can be built residue by residue on a solid support. A two-chain, multiply glycosylated heterodimer cannot: glycosylation is enzymatic and performed by a living cell, so the molecule must be extracted from a biological source or expressed in a cell system engineered to do it. Our guide on how peptides are made covers the synthetic route; glycoproteins sit outside it. And because different glycoforms have different masses and activities, the material in a vial is a population of related molecules rather than one exact compound — which is the root of the units problem below. HCG is stocked as HCG 10,000 IU; the label is a potency, not a mass.

HMG: what “combined activity” means

Human menopausal gonadotropin, also called menotropins, is a preparation rather than a molecule. DrugBank lists it (accession DB00032) as “a purified combination of human luteinizing hormone and follicle stimulating hormone.”

Its origin follows directly from the feedback loop described earlier. Gonadotropin concentrations rise after gonadal steroid output falls away, so postmenopausal urine is a naturally enriched source of both LH and FSH. Extraction methods for this were developed by Piero Donini in 1949, and the resulting preparation was introduced clinically by Bruno Lunenfeld in 1961. Early preparations contained FSH and LH in roughly a 1:1 activity ratio; later products shifted the balance and were progressively more purified, and recombinant single-hormone products have largely displaced the urine-derived material. HMG is an approved medicine in various jurisdictions; how it is used clinically is outside the scope of this guide and is not described here.

“Combined activity” is the key phrase, and it means something specific. The label does not assert that the vial contains a defined mass of two named proteins. It asserts that the contents, assayed for FSH-like activity and for LH-like activity, produce a stated quantity of each. What else is in there — co-purified urinary proteins, other glycoforms, in older preparations material contributing LH-like activity from more than one source — is not fully specified by the label. A specification written in activity terms is the honest way to describe a partially purified biological extract.

HMG is listed as HMG 75 IU and is currently out of stock.

Why IU and not milligrams

This is the most transferable idea in the guide, and it applies well beyond the HPG axis. An International Unit is a unit of biological activity, not of quantity, and it sits deliberately outside the SI system. It exists because biological preparations vary with physical form, production method and batch, so — in the standard formulation — a simple measure of mass would not suffice. Two preparations with the same biological effect contain the same number of IU even when their actual masses differ.

The unit is defined by comparison, not in the abstract. The WHO Expert Committee on Biological Standardization establishes an IU through international collaborative study: a highly purified, lyophilised international reference preparation is produced, split into precisely weighed portions, and assayed across multiple laboratories using standardised assay systems; the value of the unit is then assigned by scientific consensus to a stated quantity of that specific reference batch. Every subsequent measurement in IU is, ultimately, a measurement against that ampoule. Three things follow, and they matter at the bench.

You cannot convert IU to milligrams without preparation-specific data. The conversion factor is the specific activity of that particular material, which differs between a crude urinary extract and a recombinant product, and can differ between batches. A vial labelled in IU does not tell you what it weighs, and a reconstitution calculation built on an assumed mass concentration will be wrong.

Equal IU labels do not automatically mean interchangeable material. Two products can carry the same nominal potency and still differ in composition, glycoform distribution and impurity profile. The unit reports activity in one assay, not identity.

The contrast with a synthetic peptide is instructive. Kisspeptin-10 is a defined ten-residue sequence with one exact molecular formula, so its identity and quantity are fully captured by sequence and mass — hence milligrams. The moment a product is a heterogeneous biological preparation rather than a single defined compound, mass stops being sufficient and activity units take over. HCG at 10,000 IU and HMG at 75 IU are labelled that way for exactly this reason.

Where each molecule acts

Read down the second column and the structure becomes clear: one molecule acts at the very top of the loop, two at the bottom. They are interventions at opposite ends of a feedback system — and anything acting at the gonadal tier is, by the logic of the loop, also generating a signal that propagates back upstream.

It is worth noting what is not on this map. PT-141 is frequently discussed alongside these compounds and belongs to a completely different system — the melanocortin receptor family, not the HPG axis. Sharing a general subject area is not the same as sharing a pathway.

Bench note: glycoproteins are harder to verify

A certificate of analysis for a synthetic peptide answers a well-posed question. The sequence implies an exact monoisotopic mass, so mass spectrometry either confirms identity or does not, and reversed-phase HPLC gives a purity percentage against a defined main peak — our guide on reading peptide COAs walks through those numbers.

Neither test transfers cleanly to a glycoprotein. HCG has no single exact mass, because the glycoform population spans a range; a mass spectrum is a distribution rather than a confirmation. HPLC purity is similarly awkward, since the “impurities” may include correctly folded molecules differing only in their glycans. For a preparation like HMG, purity against a single peak is not even the relevant question — the material is defined as a mixture.

What substitutes is orthogonal: immunoassay for identity, and a potency assay calibrated against a reference standard for activity. That is slower, more expensive and less absolute than a clean mass spectrum. A COA accompanying a glycoprotein preparation is answering a different question from one accompanying a synthetic peptide, and should be read with that in mind.

What the evidence does not establish

The axis physiology described above is well established. What that physiology does not license is a long list.

It does not establish what any of these compounds do in an arbitrary experimental model. Knowing that kisspeptin signalling is required for the axis to function is a statement about the endogenous system. It is not a prediction about what a synthetic C-terminal fragment does when applied to a particular cell line, tissue preparation or animal model, where receptor density, clearance and the state of the surrounding network all vary.

Species differences are substantial and are frequently glossed over. The organisation of KNDy neurons, receptor distribution and the quantitative rules of frequency-decoding are not identical between rodents and primates. A result obtained in one is a hypothesis in the other, not a translation.

The pulsatility finding is foundational but was not a large study. Belchetz and colleagues worked with a small number of lesioned rhesus monkeys. The principle has held up and become textbook, but the specific quantitative parameters — what interval counts as “pulsatile” in a given species or physiological state — are not fixed constants and should not be treated as such.

The kisspeptin fragments are not interchangeable. Kisspeptin-54, -14, -13 and -10 share a C-terminal region and a receptor, but they differ in length and in the processing they represent. We did not verify comparative clearance or potency data for these fragments in preparing this guide, and we are not going to assert any. If a claim depends on one fragment behaving like another, that claim needs its own source.

An IU label is an assay result, not a certificate of sameness. Potency assigned against a reference standard carries assay variability, and standards are periodically replaced.

Nothing here establishes anything about use in a person. This guide deliberately does not describe administration, protocols, or outcomes in humans for any of these compounds, including the two that exist as approved medicines in some jurisdictions. That omission is intentional. These products are stocked for in-vitro research and laboratory use only, and the physiology above is presented as physiology.

Frequently asked questions

What is the HPG axis? A three-tier signalling cascade with a feedback arm. GnRH neurons in the hypothalamus release GnRH in discrete pulses into the hypophysial portal circulation; gonadotrope cells in the anterior pituitary respond by releasing LH and FSH; those gonadotropins act on the gonads; and the steroids the gonads produce feed back to suppress upstream signalling. The feedback arm is what makes it a loop rather than a one-way chain.

Why does continuous GnRH exposure shut the axis down? Because the pituitary responds to the pattern of the signal, not the total amount. In rhesus monkeys whose own GnRH output had been abolished by hypothalamic lesions, Belchetz and colleagues showed in 1978 that intermittent delivery restored gonadotropin secretion while constant infusion did not, and that switching restored animals to constant delivery caused desensitisation and down-regulation.

What is kisspeptin-10 and how does it relate to kisspeptin-54? The KISS1 gene encodes a 138-residue precursor processed into several peptides sharing a common C-terminal region: kisspeptin-54 (metastin, residues 68–121), kisspeptin-14 (108–121), kisspeptin-13 (109–121) and kisspeptin-10 (112–121). Kisspeptin-10 is the shortest and, per UniProt, the most studied. All act at KISS1R, formerly GPR54.

Why is HCG called a glycoprotein rather than a peptide? It is a non-covalent heterodimer of two chains — the 116-residue alpha chain encoded by CGA, shared with LH, FSH and TSH, and a 165-residue beta chain encoded by CGB3 that confers specificity — and both carry attached carbohydrate. That carbohydrate is part of the molecule, which is why HCG cannot be built by solid-phase synthesis.

Why are HCG and HMG measured in IU instead of milligrams? Because they are biological preparations whose activity per unit mass is not constant. The International Unit is a unit of biological activity defined by assay against a WHO international reference preparation. For a heterogeneous glycoprotein or a partially purified extract, activity is the meaningful specification and mass is not; a defined synthetic peptide such as kisspeptin-10 is the reverse case, and is sold by mass.

Are HCG, HMG or kisspeptin approved for human use? No. Everything Patriot Labs stocks, including kisspeptin-10, HCG and HMG, is sold strictly for in-vitro research and laboratory use only. None of it is offered as a drug, none is intended for human or veterinary consumption, and this guide does not describe or imply any use in a person.

All Patriot Labs products are sold strictly for in-vitro research and laboratory use only. Not for human or veterinary consumption. This guide is educational and describes peptide chemistry and published research in general terms; it is not medical advice, does not describe how to use any product, and the references cited do not constitute a product claim.