Page 1
1
The BSE Inquiry / Statement No 74
Dr Alan Dickinson (scheduled to give oral evidence 11/06/98)
STATEMENT TO THE BSE INQUIRY
By Dr Alan Dickinson
Qualifications:
In 1952 I obtained a BSC in Zoology (Hons) at Liverpool University. In
1954 I was awarded a PhD in Genetics at Birmingham University.
Research:
In 1954 I joined the staff of the Animal Breeding Research Organisation ABRO, in
Edinburgh.
In 1955 I started working with scrapie with Dr J T Stamp, who had recently been
appointed Director of Moredun Research Institute (MRI), Edinburgh. This became a long-
term collaboration.
In 1981 I became founding Director of ARC & MRC Neuropathogenesis Unit (NPU) on
the science campus of Edinburgh University. I retired early in September 1987.
My papers on TSE research are listed in Appendix 2.
1. In this statement I have sought to describe my involvement in the field of TSEs in a
chronological order. I have nevertheless found it necessary at certain points to segregate
that description into broad topics.
Personal standpoint relating to policy issues
2. Before I proceed, I feel that it is necessary to indicate my personal standpoint. Having
helped to pioneer the subject of TSEs, it is inevitable that I have described a standpoint
on various questions that I consider are directly relevant to the Inquiry as it seeks to find
the context in which an outbreak of BSE occurred, and why it turned into an epidemic. In
my view the problem stemmed from aspects of the administrative culture dominating
veterinary issues and from the progressive weakening of the autonomy of UK science
during the 1970s and 1980s. With an outbreak being fuelled by recycling the infectious
agent, speed of decision taking was crucial and this should have depended on the
judgment of the most qualified individuals, rather than based on waiting for scientific
evidence specific to BSE in cattle.
Personal standpoint relating to TSE risk assessment
3. I also wish to state my standpoint on TSE risk assessments relevant to the remit of the
Inquiry.
4. Scrapie is the prototype for TSE's and is the basis for understanding the whole group. It
is much easier to understand TSE's if analogies with conventional microbiology are
avoided. TSE diseases and causal agents are very much simpler than with other types of
infection.
5. There are many strains of these agents and they interact with the protein product of one
gene in the host animal, variants of this gene being present in all mammals. Interaction
with this gene allows the TSE agent to be replicated. When this replication takes place
in the CNS it eventually leads to fatal disease.
6. Because TSE agents do not introduce alien proteins into their hosts, no highly evolved
Page 2
2
7. A barrier undoubtedly exists to the transfer of a TSE infection from one species to
another. The `scale' of this barrier to a strain of agent passing between particular species
is unknown until tested. Where there is no absolute barrier, the risk of transfer occurring
will increase as the dose of the agent increases, so that a near-zero risk at small dose
levels can become an actual risk at massive dose levels.
8. This could be all that distinguishes BSE from scrapie, where the latter is long established
as having no significant risk for humans. It appears, however, to be widely assumed that
the single strain of agent which causes BSE is
per se
more easily transmissible to other
species than previously known strains. Such a conclusion is unsupported by critical
evidence, but it is still a possibility that has to be assumed as a worst-case option until it
is disproved.
9. Assessment of the risk of transfer of any TSE to an untested species cannot assume that
transfer between relatively closely related genera (two ruminants, for instance) is
necessarily much more likely, than transfer between very distantly related ones. The
molecular details are still uncertain of what controls the ease of transmission to new
species, but sufficient is known to exclude evolutionary proximity as the main arbiter.
10. The factor which does change at the initial replication in another species is the amino-
acid sequence of the PrP protein which is crucial to the progression of the infection this
protein plays an incompletely understood, central role in the species barrier and the
strain of agent can be an important contributory factor.
ABRO and MRI
11. ABRO was an Agricultural Research Council institute and totally core-funded as such. It
was set up in 1948 to study the extent to which classical and quantitative genetical
principles applied to livestock breeding in the wider sense. Post-doctoral staff joined
ABRO and were encouraged to develop their own interests and programmes within this
broad remit. The Moredun Research Institute to which I refer above had been set up to
work mainly on sheep diseases and had become a research institute incorporating
diagnostic and specialist adjuncts to the state veterinary service in Scotland. Since
about 1940 it has received grants from the Scottish Office Department of Agriculture and
Fisheries. Today that core funding is 66% and it gets the rest from wherever it can, for
example: Research Councils, MAFF, pharmaceutical industry, etc.
12. When I joined ABRO, it was run by a Director, Dr H P Donald, who encouraged his staff
to develop their research at their own discretion. Dr Donald was the Director until the
early 1970s.
13. In early 1955 Dr Stamp as the new Director of MRI contacted ABRO to start a
collaboration on any genetical aspects of scrapie. This is how I came to be involved in
scrapie research.
Background on scrapie research: 1935 to 1960
14. I shall now describe the state of knowledge when I joined the work on scrapie research.
All the work had to be carried out on sheep, and it was uncertain whether any of the
sheep that were to be used were already naturally infected. I never heard the agent
being referred to, then or later, as a "virus' except in occasional introductory contexts as
an "unconventional virus".
Page 3
3
15. Prior to 1955 Dr W. Gordon who had been interested in scrapie while at MRI in
Edinburgh, had left to become the Director of the Institute for Research on Animal
Diseases at Compton in Berkshire where he was initiating a scrapie programme. By
1955, Dr. D. R. Wilson at MRI was nearing retirement, having failed to make a vaccine
against scrapie but instead `only' having established a base­line for the extreme
resistance of TSE agents to inactivation by a range of chemical and physical treatments.
16. This work had been prompted by the louping-ill vaccine accident in 1935, when
approximately 10% of the recipient sheep had been infected with scrapie by an 18,000-
dose batch of the vaccine, which was made from brains of sheep infected with the
conventional virus causing louping-ill. This virus had been fully inactivated in the vaccine
by the standard procedure of adding dilute formaldehyde solution.
17. Stamp and colleagues had repeated and confirmed Wilson's work by 1959 and had
proved that the entity causing scrapie multiplied in the affected sheep and that the
disease was not simply due to the activation of a latent conventional virus by the
transmission of a non-replicating entity. There had been some ambiguous work
attempting to show that scrapie could be passed from animal to animal via pastures
somehow contaminated by the grazing of scrapie affected sheep, but there was
scepticism about this work. We had also considered and rejected another possibility,
which might falsely have given the impression of a replicating microorganism. This was
the concept of an autoimmune process which generated more of the autoantigen, as
more tissue damage occurred.
Natural scrapie in sheep.
18. In the 1950s all our research had to use sheep and had to contend with all the
uncertainties involved in using farm livestock that have figured again with BSE in cattle.
19. There had long been the general impression among sheep breeders that scrapie was
`inherited' in some manner. Using experimental mating of Suffolk sheep, we analysed
the incidence of naturally occurring scrapie, using the methods of population genetics,
and decisively excluded the then-current contention that scrapie was caused by a sheep
gene and not by an independent infective agent [1
(J/HER/20/485)
]. We further showed
that maternal transmission of the agent could play an important role in the incidence of
the disease. Our 1965 conclusions excluding gene-causation of scrapie in sheep have
recently been confirmed using molecular techniques. This work by Dr Nora Hunter and
colleagues, involving molecular analysis of the
Sip
gene in British and New Zealand
sheep, proved that the same alleles are present in both countries but scrapie does not
occur in New Zealand [Hunter, N. et al. 1997 Nature 386, 137]
(J/N/386/137).
This was
the very skilful culmination of an important initial objective when Dr Hunter joined the
NPU at the beginning of 1986.
Maternal transmission of TSEs.
20. Vertical transmission of scrapie agent, which can occur from ewe to lamb, together with
the high infectivity in sheep placentas, as a source of infection, are the likely reasons
why scrapie is an endemic sheep disease.
21. Whether maternal transmission is pre-natal, or only peri-natal, remains uncertain and its
extent is variable between different breeds and crosses [1
(J/HER/20/485)
, 20
(J/CP/84/19)
]. One of the most striking examples concerned ewes, which had been
injected subcutaneously with SSBP/1 scrapie at about the time they were mating.
Scrapie started to occur among 7% of their lambs at younger ages than previously
reported in European sheep, ranging from 7-18 months old [Gordon, W.S.1966 U.S.
Page 4
4
Dept. Agric. ARS 91-53, 19-36]. We repeated this work on a small scale, using other
breeds, and confirmed the occurrence of scrapie cases in lambs aged 8-13 months [2].
The search for different strains of scrapie agent.
22. In the mid-1950s when the question of genic causation of scrapie was being considered,
there was one obvious way forward. Could it be shown that there were different strains,
which carried, quite independently of the host sheep, the information coding for their
differences? The first hints that there might be different strains came from the work of Mr
I.H. Pattison at IRAD, during the early 1960s, in which different scrapie-passage lines in
goats showed quite different symptoms ­ `drowsy' or `scratching'.
23.
For any critical investigations it became essential to transmit scrapie to mice. In 1959
three UK workers (Chandler, Dickinson, Zlotnik) independently restarted attempts to
achieve this using a variety of scrapie agent sources. My aims were to search for genetic
controls of the disease in a wide range of mice and to search for any strains of the agent.
I found the crucial mouse gene (
Sinc
) [4
(J/CP/78/293)
, 8
(J/CP/79/363)
, 10
(J/GENR/13/213)
, 13
(J/MGG/1/2/73)
] and over a dozen agent strains were eventually
isolated from these three transmissions, most coming from the SSBP/1 source [8
(J/CP/79/363)
27
(M8 TAB 14)
, 57.
24.
During the 1960s we focussed progressively on the
Sinc
gene, which has proved to be
central to an understanding of most aspects of TSE diseases. I had realised the
importance of the gene by 1971 [13
(J/MGG/1/2/73)
], but initially we were using it as a
tool in the search for more strains of agent. It has two variants (alleles) in mice and their
difference provided the means for discovering that there is a wide range of scrapie
strains. This was based on the simple but exceptionally precise way in which strains
differ in the way they interact with these alleles ­ the result being seen as relative
differences in incubation periods in the three
Sinc
genotypes. Incubation period
increases with decreasing dose of TSE agent, therefore absolute incubation period, in a
single
Sinc
genotype, provides very limited information.
25. We were able, by a variety of means, to separate individual strains from the mixture in
SSBP/1, as well as isolating others from a variety of sheep infected naturally with
scrapie. The strains found had a great diversity of biological properties and some also
differed in physical aspects such as thermal stability. With further meticulous checking,
they represented solid evidence that strain differences were coded independently of the
host's genotype, which to me became the fundamental question, at a molecular level, in
understanding the nature of TSE agents.
26. We found that strains differed also in the stability of their `inherited' properties and
eventually detected mutation occurring under predictable conditions and at a predictable
rate. By the late 1970s, we were concluding, from an increasing range of properties, that
these were interpretable in terms of the agent possessing a replicable informational
molecule. The molecule was independent of the host's genotype, and was either a small,
untranslated nucleic acid, or else something analogous and new to biology. [13
(J/MGG/1/2/73)
, 33 (
M8 TAB 13)
, 38, 39, 48 (
J/GV/68/79)
, 51, 56
(J/GV/72/595)
, 57].
Several of these well-defined strains were widely distributed to other workers.
Page 5
5
The context of biological research and science administration affecting the ability to
respond to the outbreak of BSE.
The planning and funding of research in the 1970s
27. I consider that one element of the BSE disaster lay in the slow politicisation of the
Research Councils. This was accelerated during the mid-70's by the Rothschild Report
(
M19 TAB 2A
) which engulfed the ARC more quickly than the other Councils. The
Research Councils were deliberately set up in the late 1920's as adjuncts of the Privy
Council so as to avoid this type of encroachment by Whitehall Ministries.
28. As I understood it at the time, the political intention, from 1971, was to curtail expenditure
by the Research Councils and redirect their research choices into so-called
`consumer/contractor' priorities and leading eventually to `corporate planning' styles of
management. This was focused on the publication of the Rothschild report. Imposition
of this centralised planning-by-committees method, initially to the largest extent on the
ARC, was for some years partially held at bay by an ARC/MAFF Joint Consultative
Organisation. Even so, the threat and its implementation widely demoralised those
responsible for valuable science, and particularly basic science, while strengthening the
hand of vested interests in the Ministries.
29. In this atmosphere, short-term research appointments progressively dictated the
scientific options, priorities being decided by committees of `experts' ­ some of whom
may not have been able to meet the necessary standard for scientific decision taking.
By which I mean, sufficient familiarity with the topic they were about to decide to be able
to design, without any prompting, two small but relevant experiments on quite different
aspects of the subject. Particularly when budgets are reducing, some officials, whether
in Ministries or in Research Councils, have the opportunity to exert great power
compared with Ministers or those briefly in charge of a Research Council.
The properties of
Sinc
(details in Appendix 1)
30. By 1970, the exceptional degree of control exerted by
Sinc
(see paragraph 24 above)
on
the replication of scrapie agents, coupled with the variety of detailed ways in which it
interacted with different strains of agent, convinced me that the function of this gene was
something new to biology.
31. I concluded in 1971 that the protein product of
Sinc
acted as the replication site for TSE
agents [13
(J/MGG/1/2/73)
] and that TSEs were slow diseases because there was a
restricted number of these sites. This had some implications at the molecular level and
gave rise to several predictions. One implication, if it was the replication site, was that
its absence would prevent replication, which is what others have recently shown by using
gene ablation techniques. From the postulated shortage of replication sites it was
predicted, for example, that different strains of agent would compete for them. We were
able to prove that this occurs in many types of experiments [34
(M8 TAB 15)].
32. The protein product of
Sinc
is now called PrP. It was tempting to regard the way that it
controls the replication of TSE agents, as being similar to its normal action in an animal
uninfected by scrapie. This in turn gave rise to the speculation that TSE agents might
have originated as exceptionally rare mutants from the putative normal informational
molecule with which the gene product interacted. Even in 1971 I was so certain of its
crucial role that we proposed that it might eventually be possible to deduce the nature of
TSE agents by studying the normal molecular function of
Sinc
in an uninfected animal ­
and this approach may still be the one that provides the complete solution to the nature
of TSE agents.
Page 6
6
33. By the early 1980s it was clear to us that important peculiarities of
Sinc
were shared by
Sip
in sheep and that the two were therefore likely to be homologous genes. Shortly
afterwards molecular analysis confirmed this, and proved that homologues are present in
all mammals. It is essential to emphasise that it is not yet possible to predict details of
the disease, such as the expected incubation periods, that will occur with any untested
strain of TSE with any version of this gene in any mammal.
34. From the 1960s, most of the UK workers entertained doubts whether scrapie agents,
with such extreme physical and chemical resistance to inactivation, in comparison with
viruses, differed also in not being based on information carried in a nucleic acid core.
Information had emerged in different labs indicating that if the scrapie agent was indeed
dependent on its own nucleic acid, then this must be smaller than that in viruses, and/or
somehow much better protected. These were acceptable possibilities for what we were
finding about strain differences and
Sinc
. The very small size did not represent any
problem, in our view, because the lack of immune response by the host was consistent
with the agent not coding for any alien protein.
35. Thus the Virino Hypothesis for TSE agents was formulated, in which a replicable,
informational molecule (a small untranslated nucleic acid, or something analogous) was
bound onto the product of
Sinc
as its replication site, and that this complex may also be
strongly protected by additional host proteins from extreme physical or chemical
treatments. By 1984 we had discarded the need for such additional proteins [39].
36. The Virino Hypothesis and the detailed implications of the properties of Sinc (and its
homologues) have been given so as to indicate the level of understanding of TSE
diseases available in the UK from the early 70's . Even so, I do not regard any molecular
details of these agents, whatever the hypothesis, as having any bearing on the course of
the BSE epidemic. Some aspects of risk assessment and issues of decontamination or
of protein feed sterilisation can be adequately decided at the level of phenomenology, as
can the considerable problems in assaying the amount of infective agent in different
tissues, at different times.
The composite NPU strain-typing system, involving relative incubation periods and
brain lesion profiles.
37. Dr Hugh Fraser pushed forward vigorously the area of neuropathology from 1966.He
was joined 4 years later by Dr Moira Bruce. Both Drs Fraser and Bruce were in my
ABRO unit but were stationed at the MRI as a part of our close collaboration. By 1973,
Dr Fraser had extended the system of strain-typing into the area of neuropathology, in
which the distribution and intensity of vacuolation in the brain was integrated into a
`lesion profile' [17
(J/CP/83/29)
].
38. This parameter is fundamentally dependent on being able to compare this profile for
different agent-strains in genetically identical animals ­ our chosen lines of inbred mice ­
because the extent to which it may be controlled by unidentified host genes is unknown.
Although
Sinc
interacts with the agent-strain to determine the profile, strikingly in a few
cases, the profiles are mainly influenced by the strain of agent. This is whether or not the
agent was injected directly into the CNS. The profile is unaffected by the dose of agent
which initiated the infection, which is an attribute of considerable importance.
39. An important outcome of this study of scrapie neuropathology in the early 1970s was
their discovery that some, but not all, agent-strains produced amyloid plaques of various
types on mice. Some of these types of plaques are very similar to those seen in
Alzheimer's Disease [40
(J/GBA/1/127)
, 41
(J/NEXN/44/285)
]
Page 7
7
40. The lesion profile system for strain typing is essentially independent of that using
Sinc
-
dependent patterns of incubation periods. The two methods are complementary, adding
to the degree of discrimination which this biological system provides; thus the number of
strains that it can discriminate is potentially very large. Although the system is slow, it
provides the only thoroughly tested system for strain typing. In 1997 Dr Moira Bruce
used the system to identify the strain in nvCJD cases as being identical to that in BSE in
cattle and a variety of other species suspected to be infected with it (
Nature
389, 498-
501, 1997). Without my Unit's strain-typing methods, there would still have to be some
doubt about BSE causing nvCJD.
41. One example of the type of experiment for which lesion-profile typing is essential is in
following the events in mice given two different strains of agents. When the two are
injected at different times, a third of a mouse life-span apart in extreme examples, it can
be shown that they compete for the limited supply of replication sites. Doses, intervals,
pharmacological treatments and agent strains can be varied so as to determine which
strain will kill the mouse, or whether both will: proving which one was responsible
depends on lesion-profiling.
Preconditions for maternal transmission
42. By the early 1970s it had become clear that it was necessary to have a functional
immune system in order to infect animals with TSE agents by extraneural routes the
opposite of what would generally be expected with viral or bacterial infections. With
herbivores such as sheep or cattle, the animal has a fully mature immune system long
before birth and is thus able to be infected, even before birth, if appropriate exposure to
TSE agents occurs. In contrast humans are not fully immunologically mature at birth
and, therefore, less likely to be infectable perinatally [18
(J/N/241/536)
].
The effect of activation or depression of cell types within the immune system on TSE
pathogenesis using extraneural routes of infection.
43. The realisation that scrapie agent depended for its replication outside the CNS on cells
within the lymphoreticular system was given strong support in the early 70s by our
finding that cortisone was able to reduce the ability to infect mice with scrapie agent.
The inverse was also found by, for example, the administration of the lymphoid-cell
activating drug, PHA, close to the time of injection with scrapie agent. This discovery was
highlighted by the demonstration that a scrapie preparation, diluted to almost tenfold
beyond the end-point for it to be able to infect a control mouse, was still able to infect a
mouse given a PHA injection. [21
(J/N/249/855)
, 25
(J/L/ii/198)
, 29
(J/N/272/54)
, 31
(J/CP/88/563)
]
44. This is the general type of context for considering any possible relevance of
organophosphate treatments to the BSE epidemic.
45. Such findings opened the way for the exploration of which types of immune system cells
were involved in replication and which were not. Others had already failed to find any
involvement of conventional immune responses such as antibody formation and specific
cell-mediated responses against scrapie infection. Instead, with TSEs some cells of the
immune system can replicate the agent to a small extent, and thus have a Trojan horse
role, rather than a protective one.
46. Thus, what had started to be demonstrated was the necessity of a functional immune
system as a prerequisite for infecting an animal with scrapie by extraneural routes. All of
these drug treatments were either as a course of treatments (cortisone) or as a single
injection at about the time of injection with scrapie agent. Most of the drugs used were
Page 8
8
ineffective if their administration started more than a few hours after the scrapie injection.
In other words they seemed to be suppressing, in one case, or augmenting, in the other,
the types of cell initially involved in establishing the infection.
47. Of much more promise, if one dares to think of trying to halt an established TSE
infection, was the result of using a class of drugs called polyanions. Initially we used
Dextran sulphate 500. We were able to show that a single dose of this was effective
Page 9
9
Raising the alarm that human growth hormone might be contaminated with CJD
agent.
Helping to set NHS contamination standards
52. I warned the MRC by phone in October 1976 that growth hormone might be
contaminated with CJD agent arising from the probability that occasional infected
pituitaries could be included in a batch of glands for processing. I suggested that there
was a simple method of determining whether the protocol used for preparing the
hormone would be able to exclude, with high efficiency, CJD agent from the final
Page 10
10
peri-oscular or other abrasions might account for many, or all, of the younger BSE
cases?
Rationale for formation of the ARC & MRC Neuropathogenesis Unit. The policy and
funding background.
59. From the late 1960s, Dr Stamp and I had seen the need for joint funding of TSE work by
both the ARC and MRC. At that time the MRC was funding TSE work on a small scale in
several labs and on a larger scale in Newcastle.
60. ARC funding of scrapie research was never adequate but at least it was appropriately
long-term. As a result of the USA Department of Agriculture's scrapie eradication
programme, both MRI and IRAD received 5-10 year grants for sheep experiments under
US Public Law 480. This was an Act `to increase the consumption of US agricultural
commodities in foreign countries, to improve the foreign relations of the US, and for other
purposes'. At IRAD, at this time, it had become the practice to economise by taking
lamb crops from the scrapie-injected ewes, for sale locally for human consumption.
61. Contacts with the MRC followed during the 1960s and 70s, when multiple sclerosis and
scrapie were linked in some publications, arising from experimental work where MS
tissue suspensions had presumably been contaminated with scrapie agent, and the
suspensions then being tested in sheep. Scares based on this were repeatedly recycled
in the media for two decades. We received MS Society funds
(YB76 /10.12/ 4.1-4.12)
that supported one research post to determine whether there might be suppression of
immune reactions in the brain during scrapie infection.
62. The background to our successful attempt to found the NPU stemmed from a number of
directions. My 1976 initiative concerning the contamination of human growth hormone
with CJD agent (see paragraph [52] above) had alerted the MRC interest and they
provided some funding for us to test the preparative protocol.
63. At that time of increasing funding constraints, there were also increasing ARC doubts, as
well as mine, about the balance, appropriateness and overall fragmentation of the UK
TSE work, which still bore many marks from decisions over two decades of the Scrapie
Advisory Committee of the ARC. These funding constraints focussed a common interest
by both Research Councils.
64. I was never a member of the Scrapie Advisory Committee. As I recollect it, the MRI and
IRAD Directors attended the meetings, as did some of their staffs on an
ad hoc
basis,
but none of the outside members had experience of working with scrapie. These
members were well-meaning and tended to favour the latest ideas in vogue. They were
in no position to assess the practicability and relevance of their suggestions. In my view,
the low level of funding up to the mid-70s was far less damaging than this superficial
Advisory Committee system. The members had lacked inspired leadership and were a
case of experts in their own fields influencing a subject which they were too busy to have
time to understand, which meant that they were liable to advocate inappropriate
strategies.
65. Broad acknowledgement of these inadequcies led to the formation in the mid-1970s of a
refreshed committee, imbued with more realism and led with insight by Professor Peter
Wildy, with an approach of learning about the work and encouraging what seemed well
conceived. The Wildy Committee's remit mainly concerned scrapie research at IRAD at
Compton and in Edinburgh at MRI, in collaboration with my ABRO team: some on the
Committee had close links with MRC HQ. I was a member of the Committee and my
recollection is that the full committee may have met only once.
Page 11
11
66. One of the Committee's recommendations was that an advisory group should be set up.
The advisory group was chaired by Prof Wildy. I was a member along with several
others. The group met by visiting the laboratories involved several times in 1977 and
continued to work in 1978 and 1979. Eventually we recommended a single specialised
centre for all the UK TSE work in a report in 1979/80
(YB80 /10.14/ 1.1-1.8)
. We were
given encouragement to proceed to detailed proposals by the Secretaries of both the
ARC (Sir Ralph Riley) and the MRC (Sir James Gowans). A senior staff member of
MRC, Dr Katherine Levy, visited IRAD and Edinburgh to assess the practicalities:
Edinburgh was chosen, in part because it was MRC policy to place new (non-clinical)
Units on sites which were adjacent to a suitable university science campus.
67. At the time that the NPU proposals were approved by the MRC in 1980, it was still their
policy to fund work within appropriate ARC Institutes but the funding restrictions of the
early 1980s ended this policy (e.g. withdrawal of #187,000/year from one ARC Institute).
Their continuation of funding for the NPU was not placed in doubt by this, because it was
a joint ARC & MRC Unit, with me as its Director answerable to the Secretaries of both
Councils. There were no obligatory bureaucratic intermediaries.
68. By proposing the formation of a joint-Council Unit, I promised the ARC a reduction of
40% of their costs, to be made up by the intended MRC component. The specialised
laboratories for the NPU were to be provided economically by the conversion of a
relatively new, vacated ARC building on the Edinburgh University science campus. To
save funds, the local administration of the NPU was to be shared with an adjacent ARC
Institute, which in the event worked smoothly and efficiently.
69. The funding component of this plan is not what happened and the NPU had a severe
funding crisis from its inception in 1981. The funding ratio of 70% ARC: 30% MRC was
arranged, much more unequal than we had intended, but all the other funding and
programme proposals were approved in autumn 1980 by MRC Council.
70. The final stage was to gain ARC Council approval on 14 October 1980 (see ARC
190a/80 item 5
(YB80/10.14/1.1-1.8)
. However, 36 hours before the Council met I was
contacted with an ultimatum: either I accepted a further 30% overall cut or the whole
scheme would lapse. The basis for this did not remotely take into account the damage
this would cause to a carefully considered programme, but was purely in terms of the
`kind of figure that the Council would accept'. During the next three years, Dr Levy made
repeated unsuccessful attempts to persuade ARC to restore the cut. I learned this at a
later date. The situation was not helped by a decision during that time by the
MAFF/AFRC Priorities Board, which had superceded the previous ARC/MAFF Joint
Consultative Organisation. I understand that this Board was a body devised by MAFF to
impose an overall funding policy on the council of AFRC (there is a reference to the
Board at page 20 of the Wildy Report, see below). The decision was that AFRC funding
for work on animal diseases was to be cut by 20%. In 1986, after paying the NPU
overheads and salaries, the research funds remaining, for 23 in the science group, was
#75,000.
The science funding crisis. The AFRC.
71. The science funding crisis from the early 1980s had the effect of making it almost
impossible for a Unit supported by two Councils, to obtain outside funding. There was
nearly one exception. By 1985-6 the DoH were interested in assisting the NPU and
there was a verbal promise to support a new molecular biology post. This promise
vanished at a month's notice because of the start of the AIDS crisis.
72. At this time the feeling was widespread that the slow politicisation of the Research
Councils was leaving a smaller `pot of research funds' from which MAFF had first `call'.
Page 12
12
73. Few have been granted my opportunity to experience, as a Unit Director, two Research
Councils from the `inside'. The atmosphere for research within the MRC during the early
1980s was a breath of fresh air, compared with that in the ARC, where it was clear that
the real control lay, not with the science but with the administrative apparatus. While it
was freely reported that there was a measure of tension between the
Department/Ministry and the relevant Research Council in both cases, this was a
relationship of progressive dominance in the case of MAFF over what had then become
the AFRC.
74. This change of name accompanied a `restructuring' in 1986 of a range of smaller
Institutes and Units into a few large AFRC Institutes. This trend had been vigorously
resisted, particularly by a number of those directly involved, including myself, as being a
move in the wrong direction. It certainly was in the wrong direction for an inter-Council
Unit like the NPU. This was because it heralded the medium-term end for the NPU as it
had been conceived. It would result in the withdrawal of MRC core-funding, essential for
long term planning by those competent to argue directly, at the highest levels, for the real
research priorities. In September 1986 Professor Wildy, who was adamantly against any
loss of NPU separate identity by any form of amalgamation, asked me to prepare a
detailed confidential memorandum for him, supporting this view.
75. This structuralist philosophy enveloping the AFRC was the opposite of that of the MRC,
then under the inspired leadership of Sir James Gowans, as Secretary. The few AFRC
structuralist enthusiasts had argued that their plan would create Institutes of sufficient
size to be `critical masses'. To the MRC a vibrant University science campus was the
necessary, valuable and flexible critical mass. All of this was highlighted in the first
Corporate Plan that the MRC had to produce in 1986
in which its credo was stated on
page 1.
76. "
During its development over sixty-five years the key elements in the Council's policy
have been to recognise and back excellent individuals and to devise methods of
support best suited to achieve their aims. The history of the MRC, and of medical
research worldwide, has shown the importance of exceptional scientific leadership:
the support of gifted individuals must continue to be the first call on the Council's
resources
."
The Wildy Committee
77. At the end of 1985, the Wildy Committee was appointed by the AFRC governing Council,
where MAFF was strongly represented
ex officio
by permanent officials. It had to review
all UK R & D on livestock diseases and make recommendations. Its visits and meetings
regularly included, along with Prof Wildy, two people with experience of directing
research Institutes, Prof Biggs and Dr Martin, and Prof McFerran [DANI] and Dr Watson
[CVL], both with MAFF affiliations. Occasionally the Chief Vet Mr Rees attended and a
member of AFRC Council, Mr J. Parry. The Committee visited all the relevant UK labs
except the NPU, which was deemed to have achieved the regrouping that the Committee
aimed to achieve with other disease areas.
78.
I had had close contact with Peter Wildy over a number of years, particularly during his
major involvement in helping to set up the NPU, and when he had become Chairman of
my Director's Advisory Group. During 1986 he often confided the frustration of his
attempts to get the Wildy Committee's proposals accepted as recommendations which
could be put before the AFRC governing council. With his extensive committee
experience, he several times described this committee situation as the "worst of all",
because of the view of the clear majority was being prevented by MAFF from becoming
a committee recommendation.
Page 13
13
79. Peter Wildy, died in March 1987 and Dr Bill Martin was left to negotiate a final text with
Prof Bell the head of ADAS of which CVL was a part. The Report was presented to
AFRC Council in July 1987. It was rejected a year later. I did not see the final Report
until 1995. I had great difficulty in finding any hint of the severe criticisms, which Peter
Wildy had confided in me. He had told me that the committee had concluded, early in
1986, that the overall situation at CVL was very unsatisfactory ­ more so than at the
other labs which they visited. Peter Wildy's intention was that they ought to recommend
that CVL restrict its duties to the statutory work, and must only become involved in
research "back to back" with outside proven specialists, the latter to be in control of the
research
News of the BSE outbreak.
80. By mid-1986 Dr J. Hope, an NPU biochemist, had developed quick, efficient techniques
for diagnosing TSEs and not even requiring fresh well-preserved brain tissue (as would
be needed for brain histological diagnosis). Unknown to me, CVL staff had requested
help from NPU staff, latterly in June 1987 from Dr Hope to help them to set up some of
the newer techniques for diagnosing TSEs. Unfortunately CVL were unsuccessful in
implementing Dr Hope's recommendations. Dr Hope later told me about his action of
which I approved. I was however never formally informed that BSE existed. My first news
came in early 1987 from my colleague Dr Hugh Fraser, who had been discreetly shown
histological slides of cattle brain by one of the CVL staff.
81. I regard a scrapie origin for BSE as less likely than some possibilities, but it is difficult to
imagine, in our present state of knowledge, how finding a BSE strain in sheep in our
early strain searches would have altered the course of events. The origin of BSE is likely
to remain uncertain. There are various speculative options, all except the theory which I
describe in (h) below, involved contaminated MBM as the vector. The options are:
a. Transfer to cattle from sheep, on exceptionally rare occasions during the 1970's, of
a still undiscovered strain of scrapie identical to that causing BSE.
b. Transfer to cattle during the 1970's of a newly mutant scrapie strain which had no
opportunity to become established in sheep by natural transmission.
c. That BSE had been present in cattle on a very small scale for a number of years,
somehow having avoided recycling via MBM, perhaps until rendering standards may
have declined in the early 80's- this avoids the question of its initial UK origin.
d. Imported into the UK, after routine quarantine, in an infected but healthy bovine,
which contributed infected ( preclinical or clinical) tissues to recycling. This requires
reasonable evidence that BSE could be occurring in the country of origin.
e. As 1, 3,or 4, but with UK cattle having been made more easily infectable from the
late 1970's by some widely used husbandry treatment or procedure.
f. Imported, subject to routine quarantine, in an endemically infected exotic species
destined for a UK zoo and with the carcass sent for routine recycling.
g. Imported as some infected constituent of MBM.
Page 14
14
h. By spontaneous origin in cattle. I do not regard this particular option as impossible,
given the possible origin of TSE agents implied in the virino hypothesis. However,
an extraordinary level of rigorous evidence would be required before this
explanation could be accepted in the case of BSE, or any other TSE.
82. I regard as implausable the most widely recycled theory that regards the several
thousand BSE cases prior to mid 1989 as originating from scrapie ingestion, rather than
recycled BSE agent (Bradley, R. and Wilesmith, J.W. ( 1993) Brit.Med.Bul. 49, 932-59).
This theory rests on the unlikely assumption that the early cases of BSE were all known.
83. At the earliest stage of the BSE outbreak, when cases with symptoms suggesting a
scrapie-like disease had been found, I consider that the matter should have been
referred to the NPU for rapid confirmation of diagnosis and we should have been given
control of the necessary research and been in the forefront in advising on the actions to
take. Had we been given effective control in this sense, the first emphasis would have
been the need to make policy in terms of the judgment of those who were proven
experts, rather than waiting for the scientific proof of what had to be done. Speed of
expert decision taking is the sine qua non when an infection is being recycled. What
actually happened, as I understand from Mr Maitland Mackie, who from mid 1986 was
the relatively new Chairman of the AFRC Animals Committee, is that the control of the
BSE work had been formally allocated to CVL.
84. Authoritative advice given as early as possible, from positions of control, by the 2 or 3
UK experts with proven track records, could have reduced the extent to which the
epidemic was able to progress. Dr W. Watson, who shortly became Director of CVL,
had been included at my request on the 1985 ARC & MRC Site Visit to the NPU,
specifically so that he would become fully acquainted with the focal role and expertise of
the NPU in TSE investigations. However, in the event CVL did not take the earliest
opportunity to notify me of the BSE outbreak and discuss where control of the research
should be placed for the most expert handling. About 9 months after the emergence of
BSE, Dr Watson phoned me, with permission from Prof Biggs, but I indicated that he
should at that stage act without involving me personally, because I was then uncertain
about my immediate future. The loss of essential NPU independence (especially from
the MRC's standpoint), had been imposed despite my many detailed and vigorous
warnings for over 2 years of the potential consequences that such changes would incur
(YB 86/9.00/1.1-1.3).
85. During the prolonged early delays prior to 1988, confused signals justifying them were
being given to Ministers, judging from what they stated publicly. Once a TSE was
suspected as the cause of the earliest reported cases of this neurological disease in
cattle, proof need only have taken a few days rather than over a year.
86. It would be extremely unfair if I did not emphasise that the formal allocation of BSE work
had placed the front-line staff of CVL in an impossible position. My technical criticisms
of their work must take this fact into account: the speed with which Mr J. Wilesmith
focussed on MBM as the source of infection is particularly commendable. The NPU was
represented on MAFF's Scrapie Committee by Dr Kimberlin, and thus the formal link
between NPU and CVL. He had been in this position also in the 1970s.
87. I add that I have been impressed by the very high quality of NPU research throughout
the BSE crisis, which has been achieved despite the unsatisfactory funding and
committee-control ethos. The overall research funds were given disproportionately to
those without previous experience, so the chronically underfunded NPU received
relatively little but had their work severely hindered by grant-recipients demanding
precious materials, plus instruction in the ABC of the subject. The irony was obvious to
those whose work was being interrupted and materials wasted, that many grants were
Page 15
15
for work that was worthless. There is wide agreement that very little of value has
emerged from inexperienced labs given BSE funding. In the list of "clear results",
produced at a cost of #50M, prepared for the EU at Florence in 1996 the NPU was the
origin of most of the valuable items, produced under intolerable pressures, alongside
mainly flimsy generalities from elsewhere. We need to expose enough of the details of
the committee structures, and how they were created, to make sure such a fiasco can
never recur.
My early retirement from NPU
88. This account is not complete without mentioning a personal consequence of the turmoil
of events in the early and mid 1980s which affected the science most directly relevant to
BSE. I had been fighting very vigorously, with the active support of my NPU Advisory
Committee, against AFRC reorganisation plans, which would cripple the purpose for
which the NPU had been set up. I was being supported in various ways in this by,
among others, the then Secretary of the MRC, Sir James Gowans, by Sir William Stewart
who subsequently became AFRC Secretary and in 1990 Government Chief Scientist,
and by Prof. Peter Wildy whose death has deprived the Inquiry of crucial first-hand
evidence relevant to the initial handling of the BSE outbreak and the administrative
turmoil at that time
89. About 15 AFRC Directors were affected by the 1986 amalgamations. From Pirbright, Dr
Mahy soon joined the brain-drain. I was one of two who refused to sign new contracts
of employment relinquishing our positions as Directors of our Unit or Institute. The final
attempt to restore the independence and authority of the NPU was made by Sir James
Gowans in late June 1987 at a meeting with Prof Biggs, attended by Prof Mims (and
discussed that evening by Dr Fred Brown and Sir James). I was told that Prof Biggs
had said at that meeting that he was not prepared to release the NPU.
90. Although in 1987 we lost the immediate battle for the NPU, it is significant that in 1997,
the mid-1980s chairman of the AFRC Animals Committee, Maitland Mackie, recorded his
regrets that my warnings a decade earlier were not heeded. Instead, the battle was won
on behalf of MAFF and by those determined to 'reorganise' AFRC science policy based
on administrative structures rather than on the flair and expertise of leading scientists.
By mid-1987 it was obvious to me that the BSE outbreak was getting out of control. I had
heard of nearly 100 widely distributed cases ­ and in August 1987 I said to Maitland
Mackie "If this BSE issue is not handled properly it will destroy the meat industry". Bill
Stewart's attempt to 'hold the fort' for me with the AFRC was subverted in September
1987. As there seemed to be no rational end to the research shambles that had been
imposed, I chose to retire 2 years early at that stage, weary and sad for my NPU
colleagues, but without regrets.
Page 16
16
Appendix 1
The unique properties of
Sinc
and the molecular implications for TSE agents.
1. I was starting to realise by the late-60's that this gene was central to understanding
scrapie and that its biological properties were without parallel among any known in
animals [13]. The rapid progress that molecular techniques would make was not
anticipated. We embarked on the18-year task (which I realised would consume in the
order of 15% of our resources) of producing stocks of inbred mice congenic for the 2
alleles of
Sinc
[56
(J/GV/72/595)
], expecting that these would be essential for analysing
the normal function of the gene when techniques became available. The extremely rigid
control of the speed of incubation by
Sinc
was astonishing, considering the time scales
involved. This tight control was most easily understood in terms of its control of the
replication of TSE agents being directly at the level of the primary gene product. The
tight control also implied that this was uncomplicated by any other segregating genes in
the mice.
2. The unexpected feature was that the mechanism could not be explained simply in terms
of the protein product of one variant of the gene being qualitatively or quantitatively more
efficient than the other. With some strains of agent, the
s7
variant of
Sinc
was the more
efficient, as judged by the relative length of incubation its homozygote produced, but with
other strains the reverse applied, the
Sinc p7
homozygote having the shorter incubation
(and essentially the same was eventually found to apply to
Sip
in sheep). Furthermore,
depending on which agent-strain was involved, the incubation period in the heterozygote
was at various positions relative to the two homozygous types, either between them or
even longer than the slower homozygote.
3.
The simplified layout shown below, using A or B for the two variants of
Sinc
, illustrates
these points but does not refer to actual strains: dose and route of infection are the same
for all the examples. Different scrapie strains were distinguishable because they differed
in their relative incubation periods between the 3
Sinc
genotypes of the inbred mice.
scrapie
strain 1
AA AB BB
strain 2
AA AB BB
strain 3
AA BB AB
strain 4
AA AB
BB
strain 5
AA
BB
|
strain 6
BB AA
AB
strain 7
BB AA AB
strain 8
BB AA
AB
|<---/ /--------------------------------------
LIFESPAN ------------------------------------------------------------ |
4. Obviously, one allele could not be described as producing short incubation, and the other
one long incubation because this was reversed for some strains of agent. All except two
types of pattern have been found. The two relevant variables - dose and route - can be
varied so that clinical disease for some genotypes tends to be displaced beyond the
lifespan by increasing the incubation period, namely
[1] by lower doses, or
[2] by routes with lower efficiencies than intracerebral injection.
5.
It is reasonable to expect that this general diagram and comments apply to all
TSE's, in all species.
The details for any untested combinations are unpredictable in
our present state of knowledge. [34
(M8 Tab 13)
]
6. These basic findings posed the question of how the
Sinc
gene could be operating at the
molecular level. They prompted the proposal in 1971 [13] that the gene was producing a
protein (now called PrP) which grouped to form the replication sites for TSE agents.
Page 17
17
The reversal of gene action in homozygotes, along with the overdominance of
heterozygotes with some agent-strains, implies that we are dealing with action involving
specific 3-dimentional configurations between more than one molecule of the gene
product. In other words, dimers (or higher multimers) are involved.
7. An additional implication from the simplest interpretation of the experimental findings was
that there must be a shortage of available replication sites. This would be a limiting
factor on the rate of multiplication of the agent and would account for the slow tempo of
these diseases, as well as for the plateauing of agent concentration in peripheral organs.
If this theory is correct, it would be predicted that an increase of the number of copies of
the gene could increase the number of sites and lead (in homozygotes) to quicker
incubation. This is exactly what happened when advances in molecular biology in other
labs enabled mice to be produced with multiple copies of
Sinc.
Likewise, if
Sinc
could be
completely removed from mice, it should not be possible to infect them with scrapie ­
deletion of
Sinc
has now been achieved elsewhere and shown to have this predictable
effect.
8. Removal of a proportion of the peripheral replication sites by splenectomy, predictably
increased the incubation period, as did the presence of a gene which prevented the mice
developing a spleen [14].
9. The NPU's Replication Site hypothesis for TSE agents would also predict that if the
number of sites was the limiting factor, then there should be competition for them which
could be detectable in mixed infections when two different strains of agent were injected.
In a long series of experiments [34
(M8 Tab 15)
], most involving two intraperitoneal
injections, I showed that the injection of a `slow' strain of scrapie agent could compete
with the ability of a later-injected `fast' strain to replicate, or even to establish itself (`slow'
and `fast' are relative to the
Sinc
genotype of the mice used). It was very easy to show
this where extraneural routes of infection were used, and it was possible to decide which
of the two strains would kill the animal, or whether they both would, by altering the doses
and intervals and any immunomodulatory treatments used. This begs the question of
how we knew which strain was responsible for the lethal CNS damage ­ Dr Fraser's
lesion profile analysis was able to assess which strain was active in the brain or to
indicate that both strains were. [15, 23, 34
(M8 Tab 15)
]
10. The Virino hypothesis [35
(J/L/1/1221)
, 39; Hope, J. 1994
Ann. N.Y. Acad. Sci
., 282-9;
Farquhar, C.F. et al.; 1998
Nature
391, 345]
evolved from the Replication Site
hypothesis [34] and postulates that TSE diseases are caused by a small, replicating,
informational molecule, which is not encoded by the host and is protected by the host-
coded prote