Water for tissue culture media: what purity to set and how to check it

Water is most of the medium

Every liter of medium is mostly water. A joint FAO/IAEA technical document calls water the main component of all plant tissue culture media, and a 1995 coconut embryo study, citing an earlier textbook, puts it at 95 percent of the medium. It is also often the one ingredient with no lot number, no reading and no line on the batch record.

A 2013 bulletin from the Division of Technical Resources at the US National Institutes of Health lists what tap water carries: ions such as calcium, magnesium, sodium, chloride and sulfate, dissolved organics and gases, chlorination byproducts, and microbial byproducts, endotoxins among them, that municipal treatment does not necessarily remove.

How much of that matters to a plant culture is a separate question, and the plant evidence is species-specific. The FAO/IAEA document says tap water free of heavy metals and contaminants can replace distilled water, citing banana and ginger. In the coconut study, tap water made no difference to germination, but roots were shorter after a month, and the medium started at pH 6.2 to 6.4 before adjustment, against 4.8 to 5.0 with glass-distilled water. The lesson for a production lab is not to buy the purest water available. It is to know what is in yours, keep it consistent and write it down, so a water problem can be told apart from a recipe problem.

Distilled, deionized or reverse osmosis

The NIH bulletin is blunt that a process name, RO, DI or distilled, does not on its own define water quality. Distillation can remove a wide range of contaminants, but it is slow, so the distillate is stored, and it can be contaminated whenever the container is opened or by leaching from the container itself.

Reverse osmosis rejects bacteria, pyrogens, inorganic solids and some organics at a membrane, though dissolved gases pass more easily. It is slow too, so it feeds a storage tank, and its feed water needs pretreatment to protect the membrane from chlorine, mineral deposits and particles. Deionization exchanges ions on resin and, by the bulletin's account, is the only technology that consistently reaches top-grade resistivity. But each cartridge has a finite capacity, removes no particles, pyrogens or bacteria, does little for many organics, and can harbor microbial growth.

So most systems are a chain: activated carbon for chlorine and organics, reverse osmosis for the bulk of the work, deionization to polish, then ultraviolet light and a fine final filter. The bulletin advises sizing a system to the grade needed most often, since glassware does not need the most expensive water. Whatever the chain, the media SOP should name a measured figure, not just a process. One published banana media SOP specifies distilled water and stops there.

Resistivity and conductivity, the quick check

The quick routine check is ionic content, read as resistivity in megohm-centimeters (MΩ·cm) or conductivity in microsiemens per centimeter (µS/cm). One is the inverse of the other, so 1 MΩ·cm reads as 1 µS/cm, and both shift with temperature, so readings are reported compensated to 25 °C. The theoretical maximum is 18.2 MΩ·cm, about 0.055 µS/cm.

The NIH bulletin reproduces ASTM D1193-06, which sets a minimum of 18.0 MΩ·cm for Type I water and 1.0 MΩ·cm for Type II, and lists Type II as often suitable for preparing culture media, in a guide written for biomedical labs. ISO 3696, it notes, is scoped to water for inorganic chemical analysis. Neither is a plant tissue culture specification, so each lab sets its own figure in the media SOP.

Know what the number misses. Resistivity sees ions, not organics: the bulletin cites water holding 1,000 ppb of sucrose that still read close to 18.2 MΩ·cm. It does not see microbes. A low reading is not always bad water either, since carbon dioxide from clean air can pull 18.2 MΩ·cm water down to about 10 MΩ·cm. The bulletin adds that pH is not a useful measure of water purity. And a reading taken upstream may not match what leaves the outlet, so take the one that counts where the water is drawn.

Read the water when it is drawn for a batch, not when it was made.

Why stored water does not stay put

Purified water starts changing once it leaves the system. The NIH bulletin says dispensed water can be contaminated rapidly by air and by its container, calls deionized water hungry water, and notes that purification removes the residual disinfectant that kept tap water in check, leaving storage open to microbial growth and biofilm.

Its design for a central system is an opaque tank with a 0.2 µm air filter feeding a recirculating loop with no stagnant dead legs, with ultraviolet light at 254 nm and submicron filtration to keep microbes down, and piping shielded from light against algae.

A carboy has none of that. It is a dead leg with a lid. No source here gives a shelf life for water in a carboy, so set one in the SOP, label each carboy with its fill date, keep it closed between draws, and read the water when it is drawn for a batch. A carboy reading below what the system showed is not automatically bad, given what carbon dioxide does, but the reading on the day is the one to record.

Cartridges, membranes and sanitizing

Every stage wears. Resin has a finite capacity over a cartridge's life, filters can clog and, the NIH bulletin warns, spread contamination if they are not serviced, and a reverse osmosis membrane depends on its pretreatment. The bulletin's advice is to keep regular trend data, which lets a lab anticipate maintenance, and to follow the maker's maintenance and sanitation recommendations.

For central systems it calls for sanitization at least twice a year and sometimes quarterly, and for point-of-use polishers often quarterly or semi-annually, with final filters that may need frequent replacement. It describes heat and chemical methods and warns that system materials must be compatible with the one used, so the procedure to follow is the maker's. Its summary is worth pinning above the unit: biofilms are far easier to control through design and maintenance than to clean out of a heavily contaminated system.

Log every cartridge change, filter change and sanitization with its date and who did it, beside the readings. A resistivity line that slides for weeks and recovers after a cartridge change shows a lab its real cartridge life.

What the autoclave does and does not fix

The autoclave is a backstop for one thing. A validated cycle is there to kill living microbes in the medium, including any that came in with the water. It removes nothing dissolved: the calcium, magnesium and sodium the coconut study measured in its tap water are still there afterward, still moving the pH and adding to what the recipe says it contains, and so are dissolved organics.

Nor does it destroy endotoxin. A peer-reviewed review describes endotoxins as highly heat-stable and not destroyed under regular sterilizing conditions, needing 250 °C for more than 30 minutes or 180 °C for more than three hours, conditions a medium never sees. The FAO/IAEA note that small units may use tap water after autoclaving instead of distilling it reads the same way: the autoclave deals with what is alive in tap water, not with what is dissolved in it.

The sources here do not establish whether endotoxin at lab-water levels affects plant cultures. The NIH bulletin's tight endotoxin figures are written for cell culture and biopharmaceutical work, and no plant source here sets a limit. What endotoxin does signal is bacteria growing upstream: the review notes that bacteria grow even in nutrient-poor water, and the bulletin treats endotoxin testing as a good indicator of gram-negative bacteria. A system that grows bacteria needs sanitizing, whatever the autoclave does afterward.

Verify this

Before the water feeds a media day

Read the water where it will be drawn, temperature compensated, and check it against the figure your media SOP sets. Compare it with recent readings: a steady slide can mean a cartridge nearing its capacity while it still passes.

Confirm the last cartridge change, filter change and sanitization are inside the SOP's intervals and that the unit shows no alarm. A carboy should carry its fill date, sit inside the SOP's limit, be closed, and show no film or growth. If a check fails, use another verified source or hold the media day, and record the source and reading either way.

Put the water on the batch record

The media prep piece in this series set out a batch record for tracing a struggling run, from recipe version and lots to the autoclave run and the hold. The water belongs in it too: which system, outlet or dated carboy it came from, the resistivity or conductivity reading at the draw, temperature compensated, and the dates of the last cartridge change and sanitization.

Then the question can be asked. If batches from the same carboy, or from the days after a cartridge ran down, struggled while others did not, the water is a suspect. If the readings were steady across good runs and bad, the cause is more likely elsewhere. Neither answer is certain, but both beat guessing.

Writing the water down is a practice, not a feature to switch on. In xPlant Pro, on Pro Lab and Teams, a custom media recipe's prep notes can name the water it calls for and the reading it must meet, and recurring tasks can put cartridge changes and sanitizing on the calendar. On Hobby, Pro Lab and Teams alike, a versioned SOP can make the water check a step and SOP runs can be logged, and explant records note the media used, the thread from a struggling culture back to its batch. None of that tests the water. It helps put what a batch was made with on record.