Study of psychrophilic and psychrotolerant micro-organisms isolated in cold rooms used for pharmaceutical processing



Tim Sandle, Bio Products Laboratory, Microbiology, Dagger Lane, Elstree, Hertfordshire WD6 3BX, UK. E-mail:



To examine for psychrophilic or psychrotolerant micro-organisms in pharmaceutical cold rooms (in relation to numbers, incidents and species) and to determine, where such micro-organisms are present, whether standard microbiological environmental monitoring regimes require modification. This is presented as a case study.

Methods and results

Comparative environmental monitoring within different pharmaceutical facility cold rooms (using standard mesophilic and low temperature incubation). Data were collected over two periods, 5 years apart. The results indicated that psychrophilic micro-organisms were not present and that those micro-organisms deemed psychrotolerant, primarily pseudomonads, could be grown on standard media under mesophilic conditions.


Psychrophilic micro-organisms were not detected and those considered to be psychrotolerant were only found in low numbers. Pyschrotolerant organisms were recovered under both low temperature incubation conditions and under standard conditions (between 20 and 35°C). Further evaluation may be required, using alternative agar, and microbiologists should regularly review the species recovered to note differences between different environments.

Significance and impact of the study

The study came about from requests made by US and UK regulators concerning the risk of any extremophiles present in pharmaceutical manufacturing facilities upon product safety. Regulators expressed concerns about whether standard, and accepted, environmental monitoring regimes were capable of detecting such micro-organisms. The data provide a benchmark to support pharmaceutical manufacturers in relation to their existing monitoring programmes or as a case study with which to undertake a similar study.


Within the sterile and nonsterile pharmaceutical and medical device manufacturing sectors, the purpose of microbiological environmental monitoring is to assess the microbial quality of the cleanroom. Environmental monitoring involves the collection of samples and the subsequent examination of data relating to the numbers or incidents of micro-organisms present on surfaces, in the air and from people, and from this, an assessment is made in relation to prescribed limits and the risk of microbial transfer to a critical area or into a medicinal product. Most environmental monitoring is traditionally focused upon the examination of mesophilic micro-organisms (Bock and Koops 1992), that is, micro-organisms which will grow at temperature range between 20 and 35°C on standard, highly nutritious agar (such as soya bean casein digest medium) (Moldenhauer 2008). The reasons why monitoring regimes have conventionally selected for mesophiles are because these micro-organisms are most likely to be present, because most cleanroom contaminants are introduced into the area from people and are deposited into the air as skin is shed (Cobo and Concha 2007; Wu and Liu 2007). Furthermore, it is arguable that such micro-organisms pose the greatest risk to most types of pharmaceutical products.

The weaknesses of environmental monitoring programmes are well recognized. Such weaknesses include the imprecision of the monitoring methods used to recover micro-organisms (Boschi 2006). Furthermore, with any microbiological monitoring programme, especially those reliant upon culture-based techniques, it is accepted that such monitoring programmes will only detect a small number of the micro-organisms present in the environment. This is due to the presence of so-termed viable but not culturable organisms (VBNC) and a reflection of the many millions of species, globally, that have yet to be isolated and speciated and that do not appear within the databases of commercially available microbiological identification systems (Whitman et al. 1998; Kallmeyera et al. 2012).

In the light of these limitations, most pharmaceutical and medical device manufacturers use the environmental monitoring programme as an indicator of environmental control through providing a ‘snap shot’ of conditions at a particular time, in relation to a particular cleanroom and at a particular location within the cleanroom, and then proceed to link those discrete events together through trend analysis. The information gathered is then used to assess the behaviour of the operators, the functioning of cleanroom environmental control parameters, such as air filtration, and in relation to the efficacy disinfection practices (Sandle 2012).

Outside of the production of highly specialized medicinal products, most pharmaceutical cleanrooms, used for the preparation of tablets, creams, inhalers and so on, are operated at temperatures suitable for personnel to work in (typically 18–25°C). In addition, common to many pharmaceutical facilities, there are some areas that may become slightly warmer (such as where autoclaves are operated) and there are some areas that function as cold rooms. Cold rooms are used for product storage and conditioning and have become more widespread with the advent and growth of biotechnology products, where cold conditions are required for purification steps. To minimize the risk of contamination, cold rooms are required to be designed and operated as certifiable cleanrooms. Pharmaceutical process area cold rooms vary in their temperature of operation, with 1°C as the potential minima and the maxima, which is set anywhere between 10 and 14°C. The cold rooms to which the case study refers were operated between 2 and 8°C.

There has been a regulatory shift, as noted by the authors of this article from personal experiences with medicines inspectors, towards requesting monitoring data concerning micro-organisms with optimal growth rates that fall outside of mesophilic conditions (what might be called extremophiles: micro-organism requiring severe conditions for growth as defined by extremes of temperature, pH, chemical oxidizing agents, hypersalinity or certain types of ultraviolet light) (Lambros et al. 2003; Dutta and Paul 2012). This regulatory approach has been noted with both US Food and Drug Administration and European Medicines Agency, in relation to regulatory assessments of pharmaceutical process area cold rooms. Here, regulatory agencies have enquired about the possibility of micro-organisms that can tolerate cold conditions (psychrotolerant) or that will only grow in cold conditions (psychrophilic) being present within cold room environments.

Psychrophiles or cryophiles are extremophilic organisms that are capable of growth and reproduction in cold temperatures. Such micro-organisms can grow at temperatures lower than 15°C, and most are found in the Arctic or in the sea (typically Tmin < 0°C, to, Tmax > 15°C) (Thamdrup and Fleischer 1998). There are generally considered to be two groups of bacteria that can tolerate cold temperatures: obligate psychrophiles and facultative psychrophiles or psychrotrophs with relatively broad temperature ranges for growth (Ferroni and Kaminski 1980). Obligate psychrophiles are those organisms having a growth temperature optimum of 15°C or lower and that cannot grow in a climate beyond a maximum temperature of 20°C (Metpally and Reddy 2009). Psychrophiles are more often isolated from permanently cold habitats, whereas psychrotolerant micro-organisms tend to dominate those environments that undergo thermal fluctuations (Russell 1990).

Obligate psychrophiles are adapted to their cold environment by having largely unsaturated fatty acids in their plasma membranes. Psychrophiles possess enzymes that continue to function, albeit at a reduced rate, at temperatures at or near 0°C. Psychrophile proteins do not function at the body temperatures of warm-blooded animals, and they are unable to grow at even moderate temperatures (Todar 2008). On the basis of risks in pharmaceutical processing, given most intended routes of administration, a psychrophilic contaminant would present only a low risk to the patient.

Facultative psychrophiles can survive at temperatures as low as 0°C up through approximately 40°C. These organisms exist in much larger numbers than obligate psychrophiles. They are generally not able to grow at cold temperatures (under 15°C), although they often maintain basic functioning (Sieburth 1967). These organisms have evolved to tolerate cold conditions, where adaptation has required a vast array of sequence, structural and physiological adjustments. Nonetheless, they are not as physiologically specialized as obligate psychrophiles and are usually not found in the very coldest environments (Morita and Moyer 2004). Examples of such facultative psychrophiles are micro-organisms including Listeria monocytogenes, Vibrio marinus, Pseudomonas fluorescens and Pseudomonas maltophilia. These ‘cold tolerant’ organisms are more often referred to as psychrotolerant micro-organisms (Helke and Weyland 2004).

The basis of the regulatory concern is that such micro-organisms may not, through their adaption to cold conditions, be detected due to an inability to grow or only growing in reduced numbers, where mesophilic environmental monitoring is undertaken. Regulators have argued that under conditions of mesophilic monitoring, this leads to an underestimation of the numbers of micro-organisms present in cleanrooms designed as cold rooms. Regulators further argue, perhaps more importantly, that the consequential failure to speciate those micro-organisms that may present in cold areas means that the risk to the product cannot be fully understood (in that the microbial population in the cold area cannot be surveyed to determine whether any of the micro-organisms are objectionable in relation to the particular product and potential patient population that the product is intended for).

The counter argument to concerns about such psychrophilic micro-organisms relates to their metabolic characteristics, that is, such organisms do not pose a risk to pharmaceutical processing because they cannot grow or reproduce in the product (Jiminez 2004). A further argument is that ‘cold loving’ micro-organisms may be present, but they are found in such low numbers that do not pose a risk. From these two counter positions, it can be contended that the focus should be upon risk assessment where any recovered objectionable micro-organisms in pharmaceutical processing should be considered in relation to the question: does the particular micro-organism, at a certain level, pose a risk to the patient should it contaminated a particular product? (Price 1984; Sutton 2006).

To this question, it can be argued that although many pharmaceutical products are stored in cold areas and assuming that product ingress could occur (however, unlikely given product barrier safeguards), then the typical micro-organisms found in pharmaceutical process cleanrooms would, given the environmental conditions of cold storage (10°C or lower), either not survive for long periods or would be in stasis and would therefore not proliferate. The exception to this would be whether there was a micro-organism that was specifically cold loving and was able to reproduce under such conditions. Here, the risk is compounded if the pharmaceutical manufacturer is not aware that such micro-organisms are present (Farrington 2005). The potential risk arises, then, if there are some micro-organisms present which will only reproduce at low temperatures and would not be detected by the standard environmental monitoring programme (that is, such organisms would not grow at the incubation regime currently applied and which is designed to capture ‘mesophilic’ micro-organisms).

In response to two regulatory requests, a study was undertaken of six pharmaceutical process cold rooms used for a range of activities from storing chemicals to conditioning intermediate product, to examine for psychrotolerant and psychrophilic micro-organisms. The two regulatory requests came at different time points. The first was made in 2007 and the second came in 2012. Whilst the presentation of the first set of data to the second regulator did not satisfy the charge of potential variations in microflora over time, this at least allowed, following completion of the second study, the two sets of data, 5 years apart, to be compared.

As with any study of cleanroom micoflora, there are limitations. The first limitation relates to benchmarking, for there have been very few studies of pharmaceutical cleanroom microflora (or microbiota) published in recent years (Sandle 2011). The second limitation is that most research into cleanrooms has focused on the association with mesophilic bacteria and fungi found on human skin and has cross-referenced the human microbiome to cleanroom microbiota (Owers et al. 2004; Moissl et al. 2007; La Duc et al. 2009). There has been little reference to cleanrooms functioning outside of a 20–25°C temperature range. The third limitation is that all cleanrooms differ in design and function and with the types of products processed. Therefore, the type of cleanroom environment will have an impact upon the range of micro-organisms recovered. The fourth and final limitation is with the microbial identification method. The key variables here are the size and scope of the databases used to compare cleanroom isolates and the types of methods used for analysis (whether the method is phenotypic or genotypic and then the various technological variations of these methods).

Whilst the reader should be aware of these limitations in drawing any generalizations from the work presented in this article, the approach and general findings will be of interest in considering whether mesophilic environmental monitoring regimes are effective and whether the types of micro-organisms recovered from cold rooms present a risk to pharmaceutical processes.

Materials and methods

To examine for the presence of micro-organisms that prefer cold temperatures, a study was set up to examine representative pharmaceutical cold process areas. The pharmaceutical facility examined was used for nonsterile processing and was located in south-east England. The cold rooms were maintained at a temperature of 2–8°C, and the air within the rooms was volumetrically exchanged at a rate of twenty times per hour. Within the cold rooms, routine maintenance was performed and the rooms were subject to a standard cleaning and disinfection regime equivalent to that undertaken for other cleanrooms. All personnel entering the cold rooms wore cleanroom certified garments.

For the study, two sets of environmental monitoring samples were taken. Set one was designed to detect psychrophilic and psychrotolerant micro-organisms. With the second set, similar samples were taken, in close proximity, and incubated according to the routine mesophilic incubation regime. This was designed to provide useful information to determine whether the current environmental monitoring regime provided the best assessment of microbial levels in cold areas. For example, if micro-organisms were detected in cold areas using incubation regimes targeted at lower temperatures and similar numbers and species were also detected using the current mesophilic incubation regime, then, logically, there would be no need to alter the current environmental monitoring regime.

The aim of the study was to determine how many and what type of psychrotolerant micro-organisms were present in representative cleanrooms, based on two research questions:

  1. Is there a difference in the number of micro-organisms recovered from the psychrotolerant condition and the mesophilic condition from cold rooms?
  2. Is there a difference in the types of micro-organisms recovered from the psychrotolerant condition and the mesophilic condition from cold rooms?

To provide an element of reproducibility, six cold rooms were studied. The rooms were examined during two time periods: in 2007 and again in 2012. The cleanrooms operated at ISO 14644 class 7 (European GMP equivalent Grade C).

Six samples were taken in each room:

  1. Contact plates (cm2). Three contact plates taken: near room entrance and towards back of room; wall sample (mid-way up).
  2. Settle plates (9-cm plate, 4-h exposure). Two settle plates taken: towards left and right hand sides of room (POV room entrance)
  3. Active (volumetric) air sample (one cubic metre of air sampled). One active air sample taken: approximate room centre.

Samples were taken during three different sampling sessions on separate days when each cold room was in the operational (dynamic) state. As the rooms operated at fixed temperatures and the activities in each room were assumed to be similar, no assumption was made concerning seasonality. Although seasonal variations in temperature affect microbial recovery within different habitats (Rivkin et al. 1996), within pharmaceutical facilities, conditions are relatively standardized throughout the year due to controlled operating conditions. In total, 108 samples from the psychrophilic test and 108 samples for the mesophilic test were taken in 2007 and then again in 2012 (a total of 432 samples).

The environmental monitoring was undertaken using tryptone soya agar (TSA), which is equivalent to USP described soya bean casein digest medium (SCDM) (containing casein hydrolysate, 1·5 g; soybean digest, 0·5 g; sodium chloride, 0·5 g; agar, 1·5 g; and distilled water to make 100 ml, pH 7·3) (Khamisse et al. 2012). TSA is a highly nutritious culture medium suitable for environmental monitoring as it is designed to recover a diverse range of (primarily heterotrophic) micro-organisms. This medium was chosen because it was used for routine environmental monitoring at the facility and it is recommended by the USP. It is noted that some other studies into extremophilic micro-organisms have purposely selected for oligotrophic bacteria (organisms that can live in an environment that offers very low levels of nutrients) and have elected to use media such as the low nutrient medium Reasoner's 2A agar (R2A) (Nagarkar et al. 2001; La Duc et al. 2007).

For the incubation regime:

  1. Psychrotolerant test: 2–8°C for a minimum of 7 days.
  2. Mesophilic test: 30–35°C for a minimum of 3 days followed by 20–25°C for a minimum of 6 days.

At the end of the incubation period, samples were inspected for microbial growth. Where discernible colonies are present, these were recorded as ‘colony forming units’ (CFU). Isolates detected from the psychrophilic study were identified using a phenotypic identification system (bioMerieux API test kits: API 20NE, API 50CHB, API 50CHL, API Coryne, ID 32C and ID 32 Staph). Those micro-organisms considered to be ‘true’ psychrophiles were subject to confirmatory testing by a genotypic test (MicroSeq™ 16S rDNA Bacterial Identification System, Applied Biosystems, Foster City, CA, USA).

The results from the psychrophilic study were compared to the results obtained from the mesophilic study using a test for significance (Student's t-test). This was in order to determine whether there was a statistical difference between the numbers of micro-organisms detected, between the two conditions examined. For this, the software package North West Quality Analyst (ver. 6.1; Northwest Analytics, Portland, OR, USA) was used.


The results of the two studies are set out below. Overall, the microbial burden of the six cold rooms, when examined by both types of monitoring, was low.

Study 1 (2007)

A summary of the data indicated (see Table 1):

Table 1. Results for study 1 into psychrophilic and mesophilic micro-organisms from process area cold rooms
Number ofsamplesPsychrotolerant count (CFU) per plate (n = 108)Mesophilic count (CFU) per plate (n = 108)Number of occasionswhere psychrotolerant samples recorded highercounts than mesophilic samplesNumber of occasionswhere mesophilic samplesrecorded higher counts thanpsychrophilic samples
Number of samplesdetecting growthMeancountStandarddeviationRangeNumber of samplesdetecting growthMeancountStandard deviationRange

From these data, it was concluded that:

  1. The psychrotolerant study data were generally of a low count. A large proportion of the psychrotolerant samples collected recording zero counts, 98 of 108 (91%) samples.
  2. There were very few examples of suspected psychrotolerant samples recording higher counts.

There were three occasions where psychrophilic or psychrotolerant micro-organisms were detected in higher numbers than those of mesophilic micro-organisms. Although this did not suggest that there was a significant risk of psychrotolerant micro-organisms being present which would not be detected through the mesophilic incubation conditions, it was considered to examine this through the use of statistical test for significance. For this, the Student's t-test was selected.

Student's t-test result

The t-statistic was 2·89, at 107 degrees of freedom. At a level of significance of P = 0·05, the value of the t-distribution table, for a one-sided level of significance, was 1·671. This was less than the calculated value of t (2·89). Therefore, there was a significant difference between the results of the two data sets.

Based on the examination of the mean counts, the mesophilic test plates gave far higher counts from the cold areas.

Study 2 (2012)

A summary of the data indicated (see Table 2):

Table 2. Results for study 2 into psychrophilic and mesophilic micro-organisms from process area cold rooms
Number ofsamplesPsychrotolerant count (CFU) per plate (n = 108)Mesophilic count (CFU) per plate (n = 108)Number ofoccasions wherepsychrotolerantsamples recorded highercounts than mesophilic samplesNumber of occasionswhere mesophilic samplesrecorded higher counts thanpsychrophilic samples
Number. of samplesdetecting growthMeancountStandarddeviationRangeNumber of samplesdetecting growthMeancountStandarddeviationRange

From these data, it can be concluded that:

  1. The psychrotolerant study data were generally of a low count. A large proportion of the psychrotolerant samples collected recording zero counts, 102 of 108 (94%) samples.
  2. There were very few examples of psychrotolerant samples recording higher counts.

There were two occasions where psychrophilic or psychrotolerant micro-organisms were detected in higher numbers than those of mesophilic micro-organisms; as with the data from study 1, the data were examined for significance in relation to the numbers of micro-organisms recovered using Student's t-test for significance.

Student's t-test result

The t-statistic was 1·90, at 107 degrees of freedom. At a level of significance of P = 0·05, the value of the t-distribution table, for a one-sided level of significance, was 1·671. This was less than the calculated value of t (1·90). Therefore, there was a significant difference between the results of the two data sets.

Based on the examination of the mean counts, the mesophilic test plates gave far higher counts from the cold areas.

Consideration of psychrophilic or psychrotolerant micro-organisms

The counts, from both studies, obtained indicate that for samples incubated at low temperatures, there were few examples of higher counts being recorded than for the samples incubated at temperatures designed to encourage the growth of mesophiles. The data also indicate that the mesophilic count data gave significantly higher results.

The Student's t-test was only capable of providing an indication of quantitative difference in terms of the recovery of colony forming units on microbiological agar. One important question that remained was whether the few micro-organisms recovered from the cold area monitoring were ‘true’ psychrophilic micro-organisms or whether they were, conversely, psychrotolerant and whether different species of micro-organism were recovered from the different monitoring conditions.

The micro-organisms recovered from 2007 to 2012, which could be characterized, are shown in Table 3.

Table 3. Micro-organisms recovered from 2007 to 2012
Micro-organismNumber of isolatesrecovered (n = 17)
Bacillus cereus group1
Bacillus sphaericus/fusiformis 1
Burkholderia cepacia 2
Candida lipolytica 1
Chryseobacterium indologenes 1
Micrococcus luteus 1
Microbacterium spp.1
Pseudomonas fluorescens 8
Rhodococcus spp.1
Yersinia intermedia 1

All other isolates were uncharacterizable using the identification methods applied. The uncharacterizable colonies were not detected from samples showing the higher count under psychrophilic conditions.

Literature review has determined the growth ranges and optimum temperature for growth of the micro-organisms isolated from the monitoring to be relatively broad, as set out in Table 4 below (Russell 1990; John 1994; Mannisto 2002; Khizhnyak 2003; Dempster 2008).

Table 4. Growth ranges and optimal growth ranges for selected micro-organisms
SpeciesGrowth range (°C)Optimum growth range (°C)
Burkholderia cepacia 2–4030–35
Candida lipolytica 5–3527
Leifsonia aquaticum 7–3724–28
Rhodococcus sp.5–4030
Yersinia intermedia 2–4228–30
Pseudomonas fluorescens 4–3525–30
Bacillus species 0–3730

Therefore, the assessment of each micro-organism recovered was that each was psychrotolerant and not psychrophilic. Therefore, under the conditions of the test, no psychrophilic micro-organisms were detected. Theoretically, each of the micro-organisms could be detected from mesophilic monitoring. This was because the growth ranges indicate that each micro-organism would grow at an incubation regime across the 20–200°C to 30–350°C temperature range.


This study analysed several hundred samples, from area designated as cold areas within the production facility, and subjected these samples to:

  1. An incubation temperature designed to stimulate the growth of psychrophilic or psychrotolerant micro-organisms;
  2. An incubation temperature designed to stimulate the growth of mesophilic micro-organisms.

The reason for selecting the psychrophilic incubation conditions as 2–8°C was because the temperature was the same operational temperature as with the cold rooms and is designed to promote the growth of any micro-organisms present in these areas. The 7-day incubation time was considered sufficient to allow for any stressed or damaged cells to reproduce. With the mesophilic conditions, this incubation regime reflected the dual temperature incubation regime in use within the facility.

In relation to the detection of micro-organisms from cold area monitoring, there were a few occasions when such micro-organisms were detected (about 27% of the samples). Of the few occasions where such organisms were detected, there were five occasions, across the two studies, where micro-organisms able to tolerate the cold conditions were detected in higher numbers compared with mesophilic micro-organisms. Although this did not suggest that there was a significant risk of psychrophilic or psychrotolerant micro-organisms being present which would not be detected through the mesophilic incubation conditions, it was considered prudent to examine this through the use of statistical test for significance. For this, the Student's t-test was selected.

The statistical analysis, for both sample sets, indicated that mesophilic micro-organisms were recovered in significantly higher numbers than cold-tolerant micro-organisms. When the two sets of data are examined (2007 and 2012), both sets produced near-identical results and there was little significant variation in the data in relation to incidents or numbers of organisms detected.

For the few incidents of cold room microbial isolates recovered, these isolates were examined to see whether any of the micro-organisms were known obligate psychrophiles. Against these criteria, none were detected. Several micro-organisms that appeared to psychrotolerant were detected. Unlike the typical trend from mesophilic monitoring, where the typical flora are Gram-positive (Wu and Liu 2007), there was a mix of Gram-positive and Gram-negative micro-organisms, with a bias, in terms of numbers of different species, towards Gram-negative organisms. Of these micro-organisms detected, none, other than Pseudomonas fluorescens, is normally associated with growth in a cold area or display any resilience to the environmental stressors associated with cold conditions. With the one cold-liking micro-organism, Pseudomonas fluorescens, this bacterium will additionally grow across the mesophilic test range (20–35°C). Based on this assessment, none of the recovered micro-organisms caused concern in relation to a need to adjust the mesophilic monitoring regime.

It is possible, had an alternative or even multiculture media been used or if the isolates had been subjected to a genotypic identification (such as a measure of intracellular ATP), that different species may have been recovered. However, given that the study was sufficiently robust in terms of sample numbers, taken across time, using an identification system with a reasonably comprehensive database, then it is unlikely that the division between ‘cold loving’ and ‘cold tolerant’ species would have altered.

When the results of this study are considered back to the theoretical risks of extremophilic micro-organisms occurring within pharmaceutical processing facilities, such micro-organisms are uncommon (in this study, no obligate psychrophiles were recovered). It also remains that should such organisms be present, they cannot, for the most part, grow in most product formulations, and whether or not they should be specifically examined for should be a risk-based decision based on the presentation, formulation and administration of the pharmaceutical product.

With regard to psychrotolerant micro-organisms, which can survive in cold conditions, the study indicated that these are found in low numbers. A review of characterizable species incubated indicated that each could be recovered from a mesophilic sampling regime. However, there was some variation with the types of species and it may have been that a wider variation could have been detected had an alternative agar been used. On this basis, it would be prudent for the facility examined in this case study, to be re-examined on a regular basis to detect any changes to resident microflora within cold areas.

To revisit the earlier discussion, environmental monitoring has many limitations and it should be used foremost as a means of trend analysis. It does stand, however, based on the inference from this case study that, provided the programme is well designed, the regular characterization of microbial isolates should form part of any robust environmental monitoring programme. It should be incumbent upon the microbiologist to regularly review the microflora recovered from cleanrooms (Akers 1997). With this exercise, the rate of psychrotolerant organisms can be assessed and any adjustments to the monitoring regime be made, should this be warranted.