ORIGINAL ARTICLES

Insulin Receptors are Widely

Distributed in Human Brain and Bind
Human and Porcine Insulin with Equal
Affinity
D.F.C. Hopkins*, G. Williams
Diabetes and Endocrinology Research Group, Department of
Medicine, University of Liverpool, Liverpool, UK

Insulin receptors differing structurally from those in other tissues have been demonstrated
in brain from many species. Subtle differences in binding properties have been reported
between insulin receptors in brain and other tissues, including differences in affinity of
pig brain receptors for human and porcine insulin. Insulin binding has been demonstrated
in human cerebral cortex, but insulin binding has not been characterized in other areas
of human brain. We have studied the binding of 125I labelled human insulin, and its
displacement by unlabelled human and porcine insulin, in homogenates prepared from
human hypothalamus, cerebral cortex and cerebellum obtained post-mortem from eight
non-diabetic subjects. Specific binding was demonstrated in all brain regions studied, and
displacement curves obtained with unlabelled human and porcine insulin were identical.
By contrast, unlabelled insulin-like growth factor-1 did not significantly displace 125I
labelled human insulin over the same concentration range. We therefore conclude that
insulin receptors are widely distributed in human brain and do not differ in their affinity
for human and porcine insulin.  1997 by John Wiley & Sons, Ltd.
Diabet. Med. 14: 1044–1050 (1997)
No of Figures: 5. No of Tables: 2. No of Refs: 35
KEY WORDS human insulin; porcine insulin; insulin receptor; insulin binding; human
brain
Received 17 April 1997; accepted 1 July 1997

Introduction site for insulin, as polyclonal antisera that block receptor

binding to receptors from other tissues are less potent at
There is convincing evidence that insulin acts as a inhibiting insulin binding to brain receptors.12 In some
regulatory peptide in the mammalian brain, with functions studies, subtle differences have been described between
quite distinct from its classical role in metabolism.1,2 It the binding of various species of insulin and analogues
has been suggested that insulin may regulate processes to brain and liver. Schlüter et al.,13 studying isolated
as diverse as fetal brain growth and differentiation, and the neurones from porcine cerebral cortex, reported a greater
hypothalamic control of food intake and energy balance. affinity of the neuronal receptor for porcine insulin than
Insulin was first demonstrated in rat brain in 1978,3 for human insulin. Gammeltoft et al.,14 studying insulin
and since then insulin receptors have been demonstrated binding to rat cerebral cortex synaptosomes and liver
in the brains of diverse species.4–6 Extensive studies have showed differences between the two tissues in relative
been made in rat brain, in which receptors are widely affinities for proinsulin and coypu insulin, but observed no
distributed, with particular high densitites in the hypo- differences in the binding of porcine and human insulin.
thalamus and olfactory bulbs.7,8 These studies have By comparison with these detailed animal studies,
revealed the presence of a distinct neuronal subtype of there are relatively few data on insulin and receptors in
insulin receptor, unique to the central nervous system,9 human brain. Insulin receptors have been demonstrated
which has a lower molecular weight than insulin in human retina and cerebral cortex;15–17 as in the rat,
receptors in other tissues due to differences in glycosyl- these exhibit differences in glycosylation from receptors
ation.10,11 These differences appear to affect the binding in other tissues. One study of isolated receptors from
cerebral cortex did not find any differences in the affinity
of the receptor for human, porcine or bovine insulin.18
* Correspondence to: Dr David Hopkins, Department of Medicine, There have been no studies of insulin binding in other
King’s College School of Medicine & Dentistry, Bessemer Road, London regions of human brain.
SE5 9PJ, UK
Sponsors: Peel Medical Research Trust; Liverpool Diabetes Research In the present study, we have extended these obser-
Action Fund vations on the brain human insulin receptor by charac-
1044 CCC 0742–3071/97/141044–07$17.50
 1997 by John Wiley & Sons, Ltd. DIABETIC MEDICINE, 1997; 14: 1044–1050
 ORIGINAL ARTICLES
terizing insulin binding to membrane homogenates 0.1 M and 0.2 mM, respectively, the supernatant from
prepared from various regions of the human brain, this step was further centrifuged at 40 000 g and 4 °C
notably the hypothalamus, which appears particularly for 40 min. The resultant pellet was then washed
important as a site of insulin action in the rat. We have and re-suspended in 0.9 % saline containing 25 mM
also compared the binding of human and porcine insulin HEPES (pH 7.8).
in each of the brain areas studied with that observed in Protein content in all membrane homogenates prep-
liver, to determine if there are any local differences in aration was determined using the method of Lowry21
affinity for the two species of insulin in any of the brain with bovine serum albumin standards.
areas studied.
Binding Assay
Subjects and Methods
In preliminary experiments, insulin binding was found
Tissue Collection to be temperature- and pH-dependent; optimum binding
occurred at pH 7.8 and 4 °C, and equilibrium was
Tissue was obtained at post-mortem examination within
reached after 16 h incubation. All subsequent assays
4 to 20 h of death, from 8 subjects aged 62–90 years
were therefore performed under these conditions.
(Table 1). None of the subjects had a past history of
Samples were incubated in 96-well microtitre plates,
diabetes or neurological disease, and all brains were
in a total incubation volume of 200 ml HEPES-buffered
macroscopically normal. Samples were taken from frontal
saline containing 1 % BSA and 0.01 mg ml−1 bacitracin.
cortex, cerebellar cortex, and whole hypothalamus.
Membrane preparations were diluted to give a final
Additional samples were also obtained from occipital and
protein content of 200 mg per assay well, and this
temporal lobe from two patients. Tissue was immediately
was incubated with 0.05–0.1 pM 125I-(Tyr-A-14)-labelled
snap-frozen in liquid nitrogen and stored at −70 °C until
human insulin (Amersham, Bucks, UK) and 0.001–
processing. In addition to brain, samples of liver were
1000 nM human or porcine unlabelled insulin (Novo
obtained in four cases.
Nordisk, Crawley, West Sussex, UK). Further assays were
For comparison, samples of fresh porcine cerebral
performed using 0.001–100 nM IGF-1 (Sigma, Poole,
cortex from three animals were obtained from a local
Dorset, UK) in place of unlabelled insulin to confirm
abattoir, snap-frozen in liquid nitrogen, and stored
specificity of binding. Free hormone was separated from
as above.
bound by filtration of samples through glass-fibre mats
pre-soaked in incubation buffer, using an automated
Membrane Preparation multi-channel cell harvester, and filter-bound labelled
ligand was counted using an automated gamma-counter.
Membrane-rich homogenates were prepared using stan-
dard methods.19,20 Briefly, frozen tissue samples were
finely chopped and homogenized in 10 volumes of ice- Data Analysis
cold 0.32 M sucrose containing 0.01 mg ml−1 bacitracin,
and the crude homogenates were then centrifuged at Data from each individual binding assay were analysed
600 g and 4 °C for 10 min. For brain tissues, the resultant separately using EBDA-LIGAND software22 (Biosoft, Cam-
supernatant was further centrifuged for 20 min at 17 000 g bridge, UK), to determine values for high- and low-
and 4 °C. The pellet from this centrifugation was then affinity dissociation constants, Kd1 and Kd2, and estimates
washed and re-suspended in 0.9 % saline containing for high- and low-affinity binding capacities, Bmax1
25 mM HEPES (pH 7.8). and Bmax2.
For liver samples, the initial supernatant was centri- Student’s paired sample t-test was used to compare
fuged at 12 000 g and 4 °C for 30 min. After addition of dissociation constants obtained using human and porcine
NaCl and MgSO4 to achieve final concentrations of insulin in all tissues studied. Comparisons of dissociation

Table 1. Details of subjects used in the study

Patient number Age Sex Weight (kg) Cause of death Time to post mortem (h)

1 80 F 75 Bronchopneumonia 9
2 62 M 70 Myocardial infarction 16
3 71 M 83 Perforated peptic ulcer 20
4 71 F 68 Carcinoma of bronchus 16
5 71 M 60 Carcinoma of bronchus 6
6 62 M 74 Myocardial infarction 4
7 77 F 62 Perforated peptic ulcer 11
8 90 F 90 Bronchopneumonia 10

INSULIN RECEPTORS IN HUMAN BRAIN 1045

 1997 by John Wiley & Sons, Ltd. Diabet. Med. 14: 1044–1050 (1997)
 ORIGINAL ARTICLES
constants and binding affinities between tissues were software was obtained by assuming a two-site model,
made by one-way analysis of variance. consistent with the presence of high- and low-affinity
binding sites.
Results Mean values for the high- and low-affinity dissociation
constants, Kd1 and Kd2, and for maximum high- and low-
Binding of Human Insulin affinity insulin binding capacity, Bmax1 and Bmax2 , derived
from Scatchard analysis are summarized in Table 2.
Specific insulin binding was demonstrated in all human One-way analysis of variance revealed no significant
brain and liver membrane preparations assayed, and differences in receptor affinity between tissues. The
binding of 125I-labelled human insulin was progressively number of high-affinity binding sites (Bmax1) was signifi-
displaced by addition of increasing concentrations of cantly greater for liver than for any of the brain regions
unlabelled human insulin. Mean displacement curves (p , 0.001). Bmax1 was similar in hypothalamus and
for liver and each of the three brain areas studied are cerebellar cortex, but significantly lower in cerebral
shown in Figure 1. Maximal insulin binding to liver was cortex (p = 0.02).
considerably greater than to brain, 9.2 % of total labelled
insulin added being specifically bound to liver mem- Displacement of Binding by Porcine
branes compared with 2.4–2.8 % bound to brain prep- Insulin
arations.
Scatchard transformation of displacement data yielded Increasing concentrations of unlabelled porcine insulin
consistent curvilinear plots. A representative Scatchard also progressively displaced labelled human insulin,
plot obtained for hypothalamus is shown in Figure 2. In resulting in displacement curves that were identical to
all cases, best fit of the data by the EBDA-LIGAND those obtained with unlabelled human insulin (Figure 3).

Figure 1. Displacement of 125I human insulin binding to human hypothalamus (a), cerebral cortex (b), cerebellum (c), and liver (d)
membrane preparations by unlabelled human insulin. Figures show mean data from all tissues studied, expressed in fmol labelled
insulin bound mg−1 of tissue, adjusted for total labelled insulin of 50 fmol mg−1 tissue

1046 D.F.C. HOPKINS, G. WILLIAMS

Diabet. Med. 14: 1044–1050 (1997)  1997 by John Wiley & Sons, Ltd.
 ORIGINAL ARTICLES
Binding of IGF-1
IGF-1 did not displace 125I-human insulin significantly
until its concentration exceeded 10 nM, i.e. over 1000
times the concentration of unlabelled human insulin
required to produce equivalent displacement (Figure 4).
This excludes a significant contribution of insulin binding
to IGF-1 receptors to the total binding observed.

Insulin Binding in Porcine Brain

Specific binding of 125I-labelled human insulin was
demonstrated in porcine cerebral cortex, and was dis-
placed equally by increasing concentrations of unlabelled
human and porcine insulin (Figure 5). Scatchard analysis
Figure 2. Representative Scatchard plot for displacement of 125I- of displacement data again revealed no differences in
human insulin binding to human hypothalamus by unlabelled
human insulin
displacement constants for the two species of insulin
(Kd1 human, 0.17 + 0.04 nM; Kd2 human, 72.4 + 13.1 nM;
Kd1 porcine, 0.18 + 0.09 nM; Kd2 porcine,
46.7 + 17.9 nM).

Discussion
Before insulin was demonstrated in rat brain, the central
nervous system was thought to be independent of the
influence of insulin. It has since been established that
both insulin and its receptors are widely distributed in
mammalian brain, and that circulating insulin is capable
of crossing the blood–brain barrier, probably via receptor-
mediated active transport.23,24
The role of insulin in adult brain remains controversial,
although there is now substantial evidence that insulin
influences food intake and energy balance in rodents.
Studies using labelled 2-deoxyglucose have generally
Figure 3. Comparison of displacement of 125I-human insulin borne out the findings of classical physiological experi-
binding to human hypothalamus membrane preparation by
ments that overall brain glucose utilization is independent
unlabelled human (d) and porcine (s) insulins
of the action of insulin, but local increases in glucose
uptake in response to insulin have been observed in
Mean data obtained for displacement by human and several specific brain areas in rats, notably the ventro-
porcine insulin by Scatchard analysis are shown in medial hypothalamus (VMH) which plays a key role in
Table 3. Comparison of mean displacement constants for metabolic regulation.25,26 Recent evidence has shown
each tissue by Student’s paired sample t-test revealed that this area is particularly important in triggering the
no significant differences in either low or high affinity counterregulatory responses to neuroglycopenia,27 and
dissociation constants in any of the tissues studied it can be postulated that by increasing glucose uptake
(p . 0.1). in the VMH, insulin could lower the threshold at which

Table 2. Mean values for dissociation constants and insulin binding capacities derived from Scatchard analysis of studies of
displacement of 125I-human insulin by unlabelled human insulin

Tissue Kd1 Kd2 Bmax 1 Bmax 2

Cortex 0.14 ± 0.08 65.1 ± 17.7 3.0 ± 0.3a 159.5 ± 45.5

Cerebellum 0.24 ± 0.05 88.9 ± 24.1 6.7 ± 1.7 218.5 ± 31.0
Hypothalamus 0.21 ± 0.05 30.7 ± 4.7 6.3 ± 1.6 164.1 ± 23.9
Liver 0.07 ± 0.01 116.5 ± 61.2 13.6 ± 2.9b 214.3 ± 86.53

Kd1, dissociation constant of high-affinity binding sites (nM); Kd2, dissociation constant of low-affinity binding sites (nM); Bmax1, maximum binding
capacity high affinity sites (fmol/mg protein); Bmax2, maximum binding capacity of low affinity sites (fmol/mg protein). ap = 0.02, bp = 0.001 for
differences in mean Bmax values.

INSULIN RECEPTORS IN HUMAN BRAIN 1047

 1997 by John Wiley & Sons, Ltd. Diabet. Med. 14: 1044–1050 (1997)
 ORIGINAL ARTICLES
125
Table 3. Comparison of mean values for dissociation constants derived for displacement of I-human insulin by unlabelled
human porcine insulin

Tissue Human insulin Porcine insulin

Kd1 Kd2 Bd1 Bd2

Cortex 0.14 ± 0.08 65.1 ± 17.7 0.14 ± 0.07 56.7 ± 16.0

Cerebellum 0.24 ± 0.05 88.9 ± 24.1 0.29 ± 0.07 52.6 ± 6.3
Hypothalamus 0.21 ± 0.05 30.7 ± 4.7 0.23 ± 0.04 40.1 ± 1.9
Liver 0.07 ± 0.01 116.5 ± 61.2 0.07 ± 0.01 103.9 ± 52.1

Kd1, dissociation constant of high-affinity binding sites (nM); Kd2, dissociation constant of low-affinity binding sites (nM).

could be of considerable clinical importance if human

and animal insulins were to act differently within the
brain. Since the introduction of human insulin into
clinical practice in 1984, there have been repeated
reports of reduced awareness of hypoglycaemia following
transfer from porcine to human insulin.29–31 Although
conclusive evidence of objective changes in counterregu-
latory responses in these patients has not been
obtained,32,33 differences have been observed in visual
and auditory evoked potentials recorded during hypogly-
caemia induced with human and porcine insulin.34,35
This suggests that the two insulin species may differentially
affect sensory functions in the brain, and this could
feasibly affect the subjective awareness of hypoglycaemia.
Human and porcine insulins differ in only a single
Figure 4. Comparison of displacement of 125I-human insulin amino acid substitution at the C-terminal of the B-chain,
binding to human cerebral cortex by unlabelled human insulin but this difference results in porcine insulin being more
(d) and IGF-1 (R)
lipophilic. It has been suggested that this could result in
greater penetration of porcine insulin into the CNS,
although there is no experimental evidence to support
this hypothesis. Alternatively, the amino-acid substitution
could result in the two species of insulin exhibiting
different to affinities for the human brain receptor.
We have demonstrated specific insulin binding in
human hypothalamus and cerebellum in addition to
cerebral cortex, and thus insulin receptors appear to be
widely distributed in the human brain. Insulin binding
capacity was greater in hypothalamus and cerebellum
than in cerebral cortex, and this is broadly in keeping
with the distribution of receptors described in the rat,
although in most rat studies the density of hypothalamic
insulin binding sites has been relatively greater compared
with other brain regions. Although the distribution of
IGF-1 receptors in rat brain is similar to that of insulin
Figure 5. Comparison of displacement of 125I-human insulin receptors, the specificity of our observed results for
binding to porcine cerebral cortex membrane preparation by insulin is confirmed by the lack of displacement of 125I
unlabelled human (d) and porcine (s) insulins
insulin in the presence of excess IGF-1.
In contrast to the previous work on porcine brain,13
these responses are activated. Although direct evidence we were unable to demonstrate any differences in the
is lacking, this mechanism could explain the results of binding affinity of the human brain receptor to human
human experiments in which high-dose insulin infusion and porcine insulin, as both displaced labelled human
resulted in attenuated secretion of counterregulatory insulin equally in all brain regions studied. This is in
hormones when compared with identical hypoglycaemia keeping with the recent findings of Kotzke et al.,18 who
induced by a lower dose infusion. 28 observed similar binding of human, porcine, and bovine
Any influence of insulin on central glucose regulation insulin to receptors isolated from human cerebral cortex.
1048 D.F.C. HOPKINS, G. WILLIAMS

Diabet. Med. 14: 1044–1050 (1997)  1997 by John Wiley & Sons, Ltd.
 ORIGINAL ARTICLES
However, the affinity constants for the binding of the are widely distributed in the central nervous system of
three insulin species obtained were derived by Scatchard the rat. Nature 1978; 272: 827–829.
5. Bassas L, DePablo F, Lesniak MA, Roth J. The insulin
analysis assuming a one-site model of insulin binding, receptors of chick embryo show tissue specific structural
although a two-site model was used for defining binding differences which parallel those of the insulin-like growth
capacity. In keeping with most studies of insulin binding, factor receptors. Endocrinology 1987; 121: 1468–1476.
our Scatchard plots were all consistent with a two-site 6. Shemer J, Raizada M, LeRoith D. Structural and functional
model, indicating the presence of receptors in high- and studies on insulin receptors from alligator brain and liver.
Comp Biochem Physiol 1987; 85B: 1–10.
low-affinity states, and all the data presented in this 7. Unger J, McNeil TH, Moxley TR, White M, Moss
paper are based on this model. Consequently, direct A, Livingston JN. Distribution of insulin receptor-like
comparisons cannot be made between our data and immunoreactivity in the rat forebrain. Neuroscience 1989;
those of Kotzke, although our conclusions in respect of 31: 143–157.
the binding of human and porcine insulin to human 8. Hill JM, Lesniak MA, Pert CB, Roth J. Autoradiographic
localisation of insulin receptors in rat brain: prominence
brain are the same. in olfactory and limbic areas. Neuroscience 1986; 17:
The discrepancy with the findings of Schlüter13 may 1127–1138.
reflect methodological differences, as our assay conditions 9. Hendricks SA, Agardh CD, Taylor SI, Roth J. Unique
were markedly different from those used in this early features of the insulin receptor in rat brain. J Neurochem
study: Schlüter performed incubation for 15 min at 37 °C, 1984; 43: 1302–1309.
10. Heidenreich KA, Brandenburg D. Oligosaccharide hetero-
using a higher concentration of labelled insulin of lower
geneity of insulin receptors. Comparison of N-linked
specific activity. Under such conditions, increased insulin glycosylation of insulin receptors in adipocytes and brain.
degradation and a greater variability in steady-state Endocrinology 1986; 118: 1835–1842.
binding may be expected, which may have contributed 11. Yip CC, Moule ML, Yeung CWT. Characterisation of
to the observed results. insulin receptor subunits in brain and other tissues by
photo-affinity labelling. Biochem Biophys Res Commun
In summary, we have demonstrated that insulin recep-
1980; 96: 1671–1678.
tors are widely distributed in human brain, and bind 12. Heidenreich KA, Zanheiser NR, Berhanu P, Brandenburg
human and porcine insulin with equal affinity. We D, Olfesky JM. Structural differences between insulin
conclude that any differences in the action of human receptors in the brain and peripheral target tissues. J Biol
and porcine insulin in the human brain do not result Chem 1983; 258: 8527–8530.
from differences in receptor binding. The possibility that 13. Schlüter KJ, André J, Enzmann F, Kerp L. Insulin receptor
binding in pork brain: different affinities of porcine and
these two insulins may penetrate the central nervous
human insulin. Horm Metabol Res 1984; 16: 411–414.
system to a different extent still needs to be addressed, 14. Gammeltoft S, Staun-Olsen P, Ottesen B, Fahrenkrug J.
and the significance of the wide distribution of receptors Insulin receptors in rat brain cortex. Kinetic evidence for
in human brain needs to be determined, given the lack a receptor subtype in the central nervous system. Peptides
of insulin action on glucose utilization outside the 1984; 5: 937–944.
15. Roth RA, Morgan DO, Beaudoin J, Sara V. Purification
hypothalamus. It is unlikely that the controversy over
and characterisation of the human brain insulin receptor.
potential differences in the actions of human and porcine J Biol Chem 1986; 261: 3753–3757.
insulin will be finally settled until there is a greater 16. Potau N, Escofert MA, Martinez MC. Ontogenesis of
understanding of the normal actions of insulin in the insulin receptors in human cerebral cortex. J Endocrinol
human brain. Invest 1991; 14: 53–58.
17. Rosenzweig SA, Zetterstrom C, Benjamin A. Identification
of retinal insulin receptors using site-specific antibodies
to a carboxy-terminal peptide of the human insulin a-
Acknowledgements subunit. J Biol Chem 1990; 265: 18030–18034.
18. Kotzke G, Schütt M, Missler U, Moller DE, Fehm HL,
Klein HH. Binding of human, porcine and bovine insulin
This work was financially supported by Peel Medical
to insulin receptors from human brain, muscle and
Research Trust and the Liverpool Diabetes Research adipocytes and to expressed recombinant alternatively
Action Fund. spliced insulin receptor isoforms. Diabetologia 1995; 38:
757–763.
19. Gray EG, Whittaker VP. The isolation of nerve endings
from brain: an electron-microscopic study of cell fragments
References derived by homogenization and centrifugation. J Anat
1962; 96: 79–96.
1. Schwartz MW, Figlewicz DP, Baskin DG, Woods SC, 20. Treadway JL, Pessin JE. Receptor characterisation. In:
Porte DJ. Insulin in the brain: a hormonal regulator of Siddle K, Hutton JC, eds. Peptide hormone action: a
energy balance. Endocrine Rev 1992; 13: 387–414. practical approach. Oxford: IRL Press, 1990: 43–72.
2. Wozniak M, Rydzewski B, Baker SB, Raizada MK. The 21. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein
cellular and physiological actions of insulin in the central measurement with the folin phenol reagent. J Biol Chem
nervous system. Neurochem Int 1993; 22: 1–10. 1951; 193: 265–275.
3. Havrankova J, Schmechel D, Roth J, Brownstein M. 22. Munson PJ, Rodbard D. LIGAND: a versatile computerized
Identification of insulin in rat brain. Proc Natl Acad Sci approach for the characterisation of ligand binding
USA 1978; 75: 5737–5741. systems. Anal Biochem 1980; 107: 220–239.
4. Havrankova J, Roth J, Brownstein M. Insulin receptors 23. Pardridge WM. Receptor mediated peptide transport

INSULIN RECEPTORS IN HUMAN BRAIN 1049

 1997 by John Wiley & Sons, Ltd. Diabet. Med. 14: 1044–1050 (1997)
 ORIGINAL ARTICLES
through the blood–brain barrier. Endocrine Rev 1986; 7: diabetics transferred from beef/porcine insulin to human
314–330. insulin. Lancet 1987; ii: 382–385.
24. Wallum BJ, Taborsky GJ, Porte DJ, Figlewicz DP, Jacobson 30. Berger WG, Alhaus BV. Reduced awareness of hypogly-
L, Beard JC, et al. CSF insulin levels increase during cemia after changing from porcine to human insulin in
intravenous insulin infusions in man. J Clin Endocrinol IDDM. Diabetes Care 1987; 10: 260–261.
Metab 1987; 64: 190–194. 31. Nelleman Jørgensen L, Dejgaard A, Pramming S. Human
25. Lucignani G, Namba H, Nehlig A, Porrino LJ, Kennedy insulin and hypoglycemia: a literature survey. Diabetic
C, Sokoloff L. Effects of insulin on local cerebral glucose Med 1994; 11: 925–934.
utilisation in the rat. J Cereb Bl Fl & Metab 1987; 7: 32. Patrick AW, Bodmer CW, Tieszen KL, White MC, Williams
309–314. G. Human insulin and awareness of acute hypoglycaemic
26. Hom FG, Goodner CJ, Berrie MA. A (3H)2-deoxyglucose symptoms in IDDM. Lancet 1991; 338: 528–532.
method for comparing rates of glucose metabolism and 33. Maran A, Lomas J, Archibald H, Macdonald IA, Gale
insulin responses among rat tissues in vivo. Diabetes EAM, Amiel SA. Double blind clinical and laboratory
1984; 33: 141–152. study of hypoglycemia with human and porcine insulin
27. Borg WP, Sherwin RS, During MJ, Borg MA, Shulman in diabetic patients reporting hypoglycemia unawareness
GI. Local VMH glucopenia triggers counterregulatory after transferring to human insulin. Br Med J 1993; 306:
hormone release. Diabetes 1995; 44: 180–184. 167–171.
28. Diamond MP, Hallarmon L, Starikc-Zych K, Jones TW, 34. Kern W, Lieb K, Kerner W, Born J. Differential effects of
Connolly-Howard M, Tamborlane WV, et al. Supression human and port insulin-induced hypoglycemia on neu-
of counterregulation hormone response to hypoglycemia ronal function in humans. Diabetes 1990; 39: 1091–1098.
by insulin per se. J Clin Endocrinol Metab 1991; 72: 35. Kern W, Schlosser C, Kerner W, Pietrowsky R, Born J,
1388–1390. Fehm HL. Evidence for effects of insulin on sensory
29. Teuscher A, Berger AG. Hypoglycaemia unawareness in processing in humans. Diabetes 1994; 43: 351–356.

1050 D.F.C. HOPKINS, G. WILLIAMS

Diabet. Med. 14: 1044–1050 (1997)  1997 by John Wiley & Sons, Ltd.