Pharmaceutical Water Quality in Peptide Manufacturing: The Grade No COA Reports

Pharmaceutical water quality is the single largest uncontrolled variable behind a research peptide certificate of analysis. A peptide COA reports what is in the vial after manufacturing, but it almost never reports the grade of water used to purify, formulate and rinse the material, and the compendial limits governing that water are entirely separate from the purity assay. The result is a verifiable-looking number that rests on an unverified input.

Water is the highest-volume raw material in peptide manufacturing by a wide margin. It is the mobile phase component in preparative chromatography, the medium the peptide is dissolved in before lyophilization, the rinse used on product-contact surfaces, and often the solvent a laboratory later reconstitutes the material with. Every one of those contact points transfers whatever the water carries into the product. Yet the two pharmacopoeial grades that matter here, Purified Water and Water for Injection, are controlled by their own monographs and general chapters, and a purity percentage generated by reversed-phase chromatography is blind to nearly everything those chapters measure.

The Two Compendial Water Grades and What Separates Them

United States Pharmacopeia recognises Purified Water and Water for Injection as distinct articles. Chemically the two carry identical limits. Both are held to a Stage 1 conductivity limit of 1.3 microsiemens per centimetre at 25 degrees Celsius under General Chapter 645, and both carry a total organic carbon target limit response of 500 micrograms of carbon per litre under General Chapter 643. USP is explicit that the 500 microgram figure is a target limit response rather than a raw concentration reading, because the true limit is the instrument response to a sucrose standard corrected for the response to reagent water.

The separation between the two grades is microbiological and pyrogenic, not chemical. Water for Injection carries a bacterial endotoxin specification of less than 0.25 endotoxin units per millilitre. Purified Water carries no endotoxin specification at all. That single distinction is the entire practical difference, and it is the distinction most likely to matter to a laboratory running a cell-based assay.

Microbial control is handled differently again. Neither monograph contains a microbial requirement, a point USP addresses directly in its own published guidance, on the reasoning that a universal requirement would burden users whose application does not need it. Instead, General Chapter 1231 supplies action levels: 100 colony forming units per millilitre for Purified Water and 10 colony forming units per 100 millilitres for Water for Injection. USP states that exceeding these action levels is generally taken to represent water unfit for use, and that an out-of-specification investigation must follow. In other words, a manufacturer that declares no microbial specification is still assumed to be operating against those numbers.

Why Water Systems Fail in Ways a Purity Assay Cannot Register

A water system does not fail the way a chromatographic method fails. It fails through biofilm, and biofilm is a fundamentally different measurement problem. Planktonic organisms floating free in the water are what a sample captures. Organisms attached to pipe surfaces are usually present in far greater numbers and are the actual reservoir feeding the planktonic count. USP 1231 states this directly, which means a passing microbial result describes the shed rate of a biofilm rather than its size.

The organisms that colonise these systems are specifically adapted to them. They tolerate extreme nutrient scarcity, resist antimicrobials and biocides, and form biofilms readily. The scale of that adaptation is best illustrated by a controlled observation rather than an assertion. Pumpuang and colleagues, publishing in Transactions of the Royal Society of Tropical Medicine and Hygiene in 2011, recovered Burkholderia pseudomallei from distilled water held at 25 degrees Celsius for 16 years, measuring roughly 3.8 times ten to the seventh viable cells per millilitre against 1.4 times ten to the fifth dead cells per millilitre (volume 105, pages 598-600). Distilled water contains essentially no nutrients. The organism persisted anyway, and at high density.

Key Research Findings

• Jimenez, PDA Journal of Pharmaceutical Science and Technology, 2007, volume 61, pages 383-399: analysis of FDA recall data for 134 non-sterile pharmaceutical products from 1998 to September 2006 found 48 percent of recalls were attributable to Burkholderia cepacia, Pseudomonas species or Ralstonia pickettii. Gram-negative bacteria accounted for 60 percent of recalls against 4 percent for Gram-positive organisms. B. cepacia alone was implicated in 22 percent.
• Pumpuang et al., Transactions of the Royal Society of Tropical Medicine and Hygiene, 2011, volume 105, pages 598-600: B. pseudomallei survived 16 years in distilled water at 25 degrees Celsius, with an estimated 3.8 times ten to the seventh viable cells per millilitre remaining.
• Tavares, Kozak, Balola and Sa-Correia, Clinical Microbiology Reviews, 2020: the Burkholderia cepacia complex comprises 24 closely related species carrying a three-chromosome genome, inherent resistance to antibiotics and antiseptics, and the capacity to metabolise organic matter in oligotrophic water, in some cases using antimicrobials themselves as a carbon source.
• USP General Chapter 1231 action levels: 100 colony forming units per millilitre for Purified Water, 10 colony forming units per 100 millilitres for Water for Injection. Exceeding either triggers an out-of-specification investigation.
• USP General Chapters 645 and 643: Stage 1 conductivity limit 1.3 microsiemens per centimetre at 25 degrees Celsius, total organic carbon target limit response 500 micrograms of carbon per litre. Both limits apply equally to Purified Water and Water for Injection.
• Water for Injection monograph endotoxin limit: less than 0.25 endotoxin units per millilitre. Purified Water has no endotoxin limit.

The recall data deserve emphasis because they describe outcomes rather than laboratory conditions. Gram-negative water organisms dominate the failure record by a factor of fifteen over Gram-positive organisms, and the three genera responsible are precisely the ones associated with water systems. This is not a theoretical exposure. It is the documented modal failure of aqueous pharmaceutical manufacturing.

The Thermal and Chemical Control Measures That Sit Behind a Water Grade

Hot water sanitisation is the dominant control for Water for Injection loops. USP guidance now describes a range of 65 to 80 degrees Celsius rather than the older recommendation of at least 80 degrees. The reasoning is instructive: 80 degrees is forgiving, tolerating a 10 to 15 degree loss as heat penetrates the system by convection and conduction, while 65 degrees is unforgiving of cooler locations such as outlet valves off the main loop. USP characterises the most heat-resistant biofilm organisms colonising a water system as having a D value of roughly 5 milliseconds at 80 degrees. Temperatures materially above 80 degrees are now explicitly discouraged because they degrade gaskets and diaphragms.

Failure modes cluster in predictable places. Cold sub-loops served by shell-and-tube heat exchangers frequently do not receive enough contact time during sanitisation for the trapped chilled water to reach temperature, leaving surviving biofilm to reinoculate the loop as soon as cold operation resumes. Dead legs, particularly temporary ones created by hard-piped connections to idle equipment, never mix adequately with sanitising water and contaminate the loop continuously. A third documented failure is overwhelming a still with feedwater carrying more than 100 endotoxin units per millilitre, which occurs with poor maintenance of carbon beds and when a municipal supply switches from chloramine to straight chlorine seasonally.

The production route itself is no longer a reliable proxy for quality. The European Pharmacopoeia monograph for Water for Injections, number 0169, was revised effective 1 April 2017 to permit generation by purification processes equivalent to distillation, specifically including reverse osmosis coupled with appropriate downstream techniques. Before that revision European regulation permitted distillation only, while the United States and Japanese pharmacopoeias already allowed alternatives. A certificate stating a water grade therefore no longer implies a production method, and the two must be assessed separately.

Why This Never Reaches a Peptide Certificate of Analysis

A research peptide COA typically reports chromatographic purity, identity by mass, and sometimes water content, net peptide content or endotoxin. None of those tests interrogate the water system. Chromatographic purity is a relative area measurement against the peptide-related species eluting from the column, and a water-derived bacterial contaminant is not a peptide-related species. It does not chromatograph as an impurity peak. It is invisible to the method that generates the headline number.

Endotoxin testing comes closest, and a peptide COA carrying an endotoxin result is meaningfully stronger than one without. But an endotoxin result is a finished-product measurement at one point in time on one sampled unit. It does not establish the grade of water used upstream, and it does not detect viable organisms, which are a separate attribute measured by different methods. A lyophilised peptide cake is a poor growth medium, which is why viable organisms rarely proliferate in a sealed vial, but poor growth support is not the same as absence.

The gap widens at the point of reconstitution. Once a laboratory dissolves lyophilised material, the water chosen at that moment governs the microbial and endotoxin quality of the resulting solution regardless of how clean the powder was. A certificate describing the powder says nothing about that step. This is the practical reason water grade belongs in a sourcing conversation rather than being treated as a manufacturing detail.

What a Research Buyer Can Reasonably Ask

Three questions separate a supplier with a documented water programme from one without. First, which compendial grade is used at each stage, and specifically whether the final formulation and rinse steps use water held to an endotoxin limit or water held to no endotoxin limit at all. Second, whether the water system is monitored against the General Chapter 1231 action levels and what the sanitisation frequency and method are. Third, whether the endotoxin figure on the certificate, if one is present, is generated per batch or carried forward from an earlier lot.

None of these are exotic requests. They are standard questions in any audit of an aqueous manufacturing process, and a manufacturer with a qualified water system can answer all three quickly. A manufacturer that cannot answer them is not necessarily producing poor material, but it is producing material whose quality cannot be traced past the vial.

Maple Research Labs publishes third-party analytical certificates through independent laboratory verification and documents its analytical scope openly. For related reading on the quality attributes that sit outside a purity number, see our analysis of sterility, bioburden and microbial limits testing, the cleanroom environment in which vials are filled, and Karl Fischer determination of residual water inside the vial, which measures a different water question entirely.

Limitations and Open Questions

Two limits on the above are worth stating plainly. The recall data from Jimenez cover non-sterile and sterile pharmaceutical products broadly, not research peptides specifically, and no equivalent dataset exists for the research chemical sector. Extrapolating the proportions directly would overstate what the evidence supports; the value of the data is in establishing which organisms dominate water-borne failure, not in predicting a rate for any particular supplier.

Second, the relationship between water system microbial counts and finished lyophilised peptide quality has not been quantified in published work. The mechanistic argument is sound, since contaminated water contacting product transfers contamination and endotoxin is heat-stable and survives processes that kill the organism producing it. But the magnitude of transfer under real manufacturing conditions is not established in the literature, and any supplier claiming a precise figure for it is asserting something the published evidence does not currently support.

For research purposes only. Not for human consumption. Not for diagnostic or therapeutic use.