91av

IN THE 1920s and 1930s, the literature of microbiology began to include
reports of bacteria of immense age. Microbiologists isolated bacteria from
ancient rocks and even meteorites, which had apparently been undisturbed
for centuries and perhaps millennia – yet these environments seemed unsuitable
for microbes to multiply in. One of the most controversial reports came
from Charles Lipman, of the University of California at Berkeley, who found
100 million bacteria to the gram in the interior of a dry brick from a room
in San Luis Obispo mission in California. The room had been used as a prison,
then sealed and forgotten for a century. As the brick was dry and undisturbed,
the implication was that the individual cells had survived that long. More
startling were Lipman’s reports of bacteria in the brickwork of Peruvian
pyramids some 4800 years old, and, even more amazing, in coal, which would
be around 300 million years old.

The microbiologists responsible for such claims were careful. They sterilised
the surfaces of their brick, coal, rock or other samples, broke them open
in scrupulously clean conditions, and tested only material from the interior;
nevertheless, few of their contemporaries accepted the implications of their
findings. The US Journal of Bacteriology for 1937 was host to a gentlemanly
dispute, a model for the present day, between Lipman and two microbiologists
from Washington State University in Pullman, over Lipman’s claims regarding
coal. His critics argued that recent permeation of bacteria into the samples
would explain his findings. His other claims were not overtly criticised,
but there may have been comparable circumstantial explanations. For example,
the bricks from San Luis Obispo, which contained organic matter, might well
have been damp enough on occasions for microcosms of successive bacterial
generations to have grown and died in situ almost indefinitely.

In other cases, even the precautions of experienced microbiologists
might not have been sufficient to exclude extraneous bacteria at all stages
of the operations. Microbes can multiply on imperceptible traces of organic
matter; they can survive, and even thrive, in minute films of water on particles;
and they are carried about in the air in tiny droplets. Bacteria penetrate
the microscopic pores in rocks and sediments if these are damp, carried
by capillary flow or actively swimming through the channels. Some bacteria,
such as Desulfovibrio, could travel between 0.5 and 13 metres a year in
a deep oil formation, according to Claude ZoBell of the Scripps Institution
of Oceanography in La Jolla, California.

As microbiologists became aware of how versatile and ubiquitous bacteria
are, they tended to dismiss reports such as Lipman’s as instances of contamination
by contemporary microbes. Yet some remained uneasy, especially about such
paradoxes as the presence of live bacteria in Arctic permafrost. Admittedly,
these exist in very small numbers, but they are present at sites which may
have been frozen for several centuries. Some of them, bafflingly, are thermophilic:
able to grow and multiply only at temperatures above 30 Degree C to 35 Degree
C, and doing best at about 65 Degree C.

Why, then, were – and are – microbiologists so sceptical? Because ordinary
bacteria die very easily. They are killed by quite minor stresses, such
as chilling, freezing, small doses of antibiotics, or disinfectants. Even
an abrupt change in the salinity of their environment can kill many species.
That vegetative bacteria should survive for more than a few weeks in conditions
unsuitable for multiplication is most implausible. However, bacteria can
go into states of suspended animation – cryptobiosis – in which they show
no, or extremely little, metabolic activity although they remain capable
of resuming their lives when circumstances change. They have three ways
of doing this.

One way is to form spores. Some bacteria can sense incipient stress.
They respond by undergoing a sequence of developmental changes whereby the
vegetative cell transforms itself into a relatively dry dormant body covered
with a tough coat. This spore is spherical or ovoid no matter what the original
shape of the microbe, and all metabolic activity seems to be at a standstill.
The ability to form spores is not very common among bacteria: it has not
been found among the Archaebacteria, and among the Eubacteria it is rare
outside the genera Bacillus, Clostridium and Thermoactinomyces. Where spores
do form, they are very resistant to freezing, drying, chemicals and heat;
some even survive brief autoclaving, and spores of the most heat-resistant
type, Bacillus stearothermophilus, are used routinely to check the effectiveness
of hospital autoclaves.

Some bacteria form comparable bodies which, although more resistant
to maltreatment than the vegetative cells, rarely approach spores in toughness.
Those of cyanobacteria (blue-green algae) are called akinetes, those of
Azotobacter are called cysts.

The other two means of becoming dormant are passive processes, outside
the microbes’ control. Freezing or drying kills most vegetative bacteria
but, in certain environments, some or all of the cells may retain the ability
to resume their lives if and when they are thawed out or rehydrated as the
case may be.

The best environment for preserving bacteria in the frozen state is
a solution containing 10 to 30 per cent of glycerol. (Glycerol protects
the cells of higher organisms from freezing damage, too, and is widely used
to preserve organs in the frozen state for surgical transplants.) With bacteria,
a wide variety of other solutes act as ‘cryoprotectants’, as they are called,
including sugars, solvents, detergents and even polymers .

Some cryoprotectants work at very low concentrations. When bacteria
are killed by freezing, without a cryo protectant, most of the cells burst
open, releasing their contents into the environment. Among those contents
are proteins and sugars, which can be cryoprotectants. So, provided the
population is dense enough for the natural cryoprotectants to reach an effective
concentration, the first ones to die protect some of the survivors.

Rather similar considerations apply to drying bacteria. Again, the bacteria
can remain viable if the environment contains protective agents, and sugars
are good protectants (especially when mixed with protein or another polymer).
A mixture of glucose and sterilised serum is especially versatile in protecting
a wide variety of bacteria and preserves the stocks of culture collections
in a freeze-dried condition. (The mixture is called mist. desiccans; an
archaic name with appropriate alchemical overtones because the reason why
it works so well is still something of a mystery.) But dormant, dry bacteria
can be prepared without freezing; the important prerequisites are the protective
agents. As with freezing, internal contents are released from the cells
when vegetative bacteria dry out and, in suitably dense populations, material
released from those which died first can protect their surviving fellows.

In nature, bacterial populations may well protect each other – I do
not know of any study of this question – but the natural environment is
certainly well provided with agents that have protective actions, such as
carbohydrates, proteins, other soluble polymers and oligomers. Soil, for
example, protects bacteria well against both freezing and drying. The early
soil microbiologists – and some contemporary ones – used this knowledge
to store specimens of their isolates. They would add a few drops of a culture
to sterile, dry soil, allow it to soak in and store it in a cupboard. Often
the bacteria survived there for decades. Bodily excretions and exudates
from animals and decomposing organic matter, for instance, yield protectants.
Reflect, if you will, on what happens as the nasal mucosa, with its flourishing
population of bacteria, dries on your pocket handkerchief. Happily, you
and your own nasal micrococci usually get on well with each other.

How long do such dormant bacteria remain viable? The question has not
been studied long enough to give a clear answer. In the frozen state, the
general pattern seems to be that when a population of vegetative bacteria
is frozen in the presence of a cryoprotectant, almost all will survive at
first but then the population dies slowly: with a good cryoprotectant, such
as glycerol, very slowly indeed – but quite rapidly with a less good one
such as i-erythritol . Lower temperatures favour longer survival. Dried
or freeze-dried populations seem to lose viability slowly, too. In both
cases there is anecdotal evidence that, with some species, a few of the
original bacteria survive in the frozen or dry state for more than four
decades. The rates at which viability declines suggests that this probably
represents the limit of survival, but perhaps a few out of an initial billion
bacteria might last a century or so.

With spores the situation is somewhat different. In 1962, Peter Sneath,
then working at the National Institute for Medical Research at Mill Hill,
London, already knew that a suspension of spores of the anthrax bacillus
prepared by Louis Pasteur in 1888 was still viable in 1956, and that a 118-year-old
can of meat had been found to contain spores of a thermophilic microbe.
He was also aware that ‘coliform bacilli’ had recently been found in frozen
faeces deposited by Scott’s ponies on his Antarctic expedition in 1911.
(He himself had examined a coprolite approximately 3500 years old from Mexico
and found it sterile, however.) These were isolated examples, so, was a
systematic approach to the longevity of dormant bacteria feasible?

In an admirable piece of lateral thinking, Sneath realised that in the
herbarium at Kew Gardens, in London, there existed a collection of dried
and individually packaged plant specimens which had been added to regularly
since 1640. The soil associated with these specimens should contain the
spores of bacilli of a wide range of ages. They did. Sneath found viable
bacillus spores, in decreasing numbers, in samples up to 320 years old.
Plotting their numbers, he constructed a survival curve and, from published
data on numbers of bacilli in ordinary soil, he predicted that a tonne of
dried soil would retain a few viable Bacillus spores even after 1000 years.

This experiment provided something for other microbiologists to go on,
because it had only one tiny flaw: the soils on the plant specimens would
have had different populations of bacilli when fresh – but the differences
would have been statistically trivial. Interest in cryptobiotic bacteria
began to revive. In 1963, Heinz Dombrowski of the University of Giessen
in West Germany claimed to have found viable Bacillus circulans in samples
of salt from Devonian, Silurian and Permian deposits, originating in Germany
and North America. This would have made them more than 600 million years
old. Dombrowski’s findings were immediately disputed for, although this
microbe could hardly have multiplied in salt, its spores are common in air.

The temple of the immortal microbe

Another surprising report came in 1966 from Egypt, from Yusef Abd-El-Malek
and his colleague Y. Z. Ishac, at the University of Cairo. They found viable
bacteria (clostridia and azotobacters) in the interior of a mud brick from
the Great Temple of Amon at Karnak, in Egypt. The brick came from a height
of 15 metres and was protected from rain. It was at least 2400 years old
yet the straw which had been used to make it had still not rotted. Azotobacters
were few, only 11 to 17 A. chroococcum per gram of brick dust, yet such
longevity among any of these microbes at all was hard to credit, because
azotobacters do not form spores but only the much more fragile cysts. Bacteria
belonging to this group are also notoriously sensitive to drying, though
they survive that stress best in soil. However, the soil of the Nile Delta
is notable for the highest recorded densities of azotobacters in the world,
so it is difficult not to conclude that modern bacteria, on dust from the
delta perhaps, crept into the experiment by some unintended fault of design
or manipulation.

In the mid-1970s, Tom Cross and his colleagues at the University of
Bradford found viable spores of a group of thermophiles called Thermoactinomyces
in debris beneath the Roman fort of Vindolanda, just south of Hadrian’s
Wall, near Carlisle. The deposits date from around AD 90. The researchers
also found spores of Thermoactinomyces in lake sediments in East Anglia
and the Lake District, of dates ranging from about 600 BC to about AD 50.
Soon after, Nancy Parduhn and John Watterson of the US Geological Survey
found spores of Thermoactinomyces vulgaris in stratified sediments in Elk
Lake, Minnesota. Spores were present in substantial numbers (some 1000 per
gram) in a layer of sediment about 5150 years old, with smaller numbers
in younger and older strata. The oldest stratum approached 7200 years old.

There are three possible ways by which modern spores might reach ancient
sediments or soil and so delude microbiologists. The spores, or viable fragments
of their vegetative forms which had grown elsewhere, could be carried in
by water permeating the sediment. They might also be transported by worms
and other burrowing fauna. And, finally, they might have grown there. The
researchers at Elk Lake excluded both means of transport, arguing that animals
moving through the sediment would have disturbed the strata, which were
clearly undisturbed, and that physical permeation by water could not produce
a peak in density at 5150 years old with a trough on either side.

Both Cross’s group and Parduhn and Watterson excluded growth in situ
on the grounds that Thermoactinomyces, being a thermophile, cannot grow
in cold soil or sediment. Of course, a devil’s advocate might contend that,
though thermophiles do not multiply at a temperature below 35 Degree C in
laboratory tests lasting weeks or even months, who really knows that they
do not metabolise and even multiply, albeit imperceptibly slowly, over several
centuries out there in the cold? An earlier report by J. W. Bartholomew
and George Paik, of the University of Southern California, Los Angeles,
had already taken care of this last point. In 1965, Bartholomew and Paik
were studying the floor of the Pacific Ocean off Mexico and discovered viable
spores of Bacillus stearothermophilus in cores of marine sediments at a
constant 4 Degree C. Carbon dating established that these cores were between
6000 and 8000 years old. The Americans’ incredulity was reflected in their
almost hesitant publication. As well as being a thermophile, B. stearothermophilus
is not a marine organism, so it could not have multiplied in situ because
the sediments were too salty as well as too cold. In this case, however,
transport was not rigidly excluded.

One can niggle about details of the work on sediments, but taken as
a whole the conclusion is compelling that the spores settled there as the
sediments were forming, deposited from the air or from water running off
the land in the marine case, and have survived there ever since: for up
to 8000 years.

If spores can live so long, why do they die at all? At Mill Hill, Sneath
reflected on this problem and noticed that the decay curve for his bacilli
from Kew was not unlike the decay curve of a mixture of radioactive isotopes
of different half-lives. All living material is subject to natural ionising
radiation, from such sources as cosmic rays and radiopotassium: might the
slow death of spores be caused by background radiation? At Oxford a few
years ago, two professors of microbiology, Howard Gest and Joel Mandelstam,
decided to look into the matter. Ionising radiation kills most of the bacteria
in a population, but some survive and among these are organisms which prove
to be genetic mutants. If suspensions of spores die because they are affected
by natural radioactivity, then old spore suspensions ought to contain more
mutants than fresh ones. Gest and Mandelstam had a 16-year-old suspension
of spores of Bacillus subtilis, so they compared the number of mutants in
that population with the number in a similar but fresh suspension. They
chose to look for a class of mutants called auxotrophs, in which the microbe
develops the need for a growth factor in order to multiply, because these
are easy to detect in the laboratory.

The two populations proved to be essentially identical: in each case
about five spores per thousand were auxotrophic. This is almost exactly
the frequency of ‘spontaneous’ mutation that microbial geneticists find
for their workhorse, Escherichia coli, when it is not exposed to radiation
or otherwise treated to induce mutations. This implies that background radiation
has no detectable influence on the viability of Bacillus spores. So there
seems to be no good reason why they should die at all. Perhaps they do not.
Sneath’s organisms at Kew, for example, may, now and again over the centuries,
have become sufficiently damp and warm, on humid summer’s days perhaps,
to initiate germination – only to die of starvation. Spores deep in soil
or in sediments are generally screened from such fluctuations, except over
geological time, when glaciations, warmings or tectonic activity might have
relevant effects.

There may be much older old spores out there, waiting for energetic
microbiologists to revive them. As Gest and Mandelstam point out, the time
limits of microbial cryptobiosis have become an open question again.

* * *

How to protect bacteria from freezing to death

FREEZING kills most vegetative bacteria but, like the cells of higher
organisms, they can be protected from damage by cryoprotectants such as
glycerol or dimethylsulphoxide. Concentrations of 10 to 50 per cent glycerol
are used routinely to store organs or microbes in the frozen state. Lower
concentrations work with bacteria, 3 per cent being about the minimum.

Using Klebsiella aerogenes as a test organism, a range of other substances
has been shown to protect bacteria from freezing damage. The following are
examples of substances that give good or complete protection at 10 per cent
concentration:

Polyalcohols: glycerol, diethylene glycol, i-erythritol

Sugars: glucose, sucrose

Sulphoxide: dimethylsulphoxide

Amides: dimethylacetamide, dimethylformide, acetamide

Soluble polymers: polyvinyl pyrrolidone, polyethylene glycol

Soluble proteins: human albumin, ovalbumin

Mixture: a trypsin digest of meat

Detergents (at 3 per cent): ‘Tween 80’, ‘Macrocyclon’, ‘Triton WR1339’;
these substances are special in being effective above 0.5 per cent.

Once frozen, the shelf-life of the population of bacteria depends very
much on the chemical nature of the cryoprotectant, as is illustrated by
the table. These data were obtained by thawing samples of a standardised
population of K. aerogenes frozen in 10 per cent solutions and assessing
their viability. Zero time means that the sample was thawed without storage
immediately after it had been frozen.

The reasons for the large differences between, for example, glycerol
and i-erythritol are not clear but are thought to involve the readiness
with which changes from the vitrified to the crystallised state take place
after freezing.

—â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä“
Protectant Store life at -20 degreesC —â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä“
Glycerol Days 0 2 6 10 20 27 40 Viability
95 92 91 86 92 85 85 Polyethylene Days 0 0.7
1.7 4 16 glycol Viability 96 80 87 70 65 Diethylene
Days 0 2 16 glycol Viability 98 72 53 —â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä“
Sucrose Hours 0 2 5 8 11.7 24 Viability
98 78 72 60 25 25 Glucose Hours 0 2 4.5
7.5 16.5 Viability 92 31 22 19 12 —â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä“
i-erythritol Minutes 0 5 10 15 20 100 Viabiluty
94 98 9 3 4 1 —â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä”â¶Ä“

John Postgate FRS is emeritus professor of microbiology at the University
of Sussex.