a r c h s o c e s p o f t a l m o l .

2 0 1 6;9 1(4):170–176

ARCHIVOS DE LA SOCIEDAD
ESPAÑOLA DE OFTALMOLOGÍA

www.elsevier.es/oftalmologia

Original article

Streptozotocin-induced diabetes, and the optic

nerve blood barrier夽

R. Alemán ∗ , B. Mompeó, I. Castaño

Departamento de Morfología, Universidad de Las Palmas de Gran Canaria, Las Palmas, Spain

a r t i c l e i n f o a b s t r a c t

Article history: Objective: To study the features of the endoneurial micro-vessels of the optic nerve in
Received 5 December 2015 streptozotocin-induced diabetic animals.
Accepted 26 December 2015 Methods: Optic nerves from control and streptozotocin-induced diabetic animals were stud-
Available online 28 March 2016 ied by light and transmission electron microscopy. Patency was determined by indirect
immunofluorescence albumin detection. The expression of major histocompatibility com-
Keywords: plex class II molecules was performed by direct immunofluorescence. The endoneurial
Optic nerve vessels were counted, and the endothelial cell, the basement membrane, and the surface of
Diabetes the transverse section of the nerve were measured.
Hyperglycaemia Results: Vessels of diabetic rats showed vessel wall thickening, preservation of pericytes,
Patency an increase in endothelial cell transcytosis, and an increased number of perivascular
Vessels macrophage cells. It may be concluded that the effects of hyperglycaemia on the inner ves-
sels of the optic nerve are more similar to the cerebral diabetic vessels than to the retinal
vessels in diabetic animals.
© 2016 Sociedad Española de Oftalmología. Published by Elsevier España, S.L.U. All rights
reserved.

Diabetes inducida por estreptozotocina y comportamiento de la barrera

sanguínea del nervio óptico

r e s u m e n

Palabras clave: Objetivos: Conocer las características de la barrera sanguínea del nervio óptico de animales
Nervio óptico con diabetes inducida por estreptozotocina.
Diabetes Método: Los nervios ópticos de animales diabéticos y controles se estudiaron mediante
Hiperglucemia microscopia óptica y microscopia electrónica de transmisión. La permeabilidad de los vasos
Permeabilidad fue determinada mediante la detección de albúmina con inmunofluorescencia indirecta y
Vasos la expresión de las moléculas del complejo mayor de histocompatibilidad clase II mediante
inmunofluorescencia directa. Asimismo, se realizó un análisis morfométrico de la superficie
del nervio, el número de vasos y el engrosamiento de la célula endotelial y lámina basal.

夽
Please cite this article as: Alemán R, Mompeó B, Castaño I. Diabetes inducida por estreptozotocina y comportamiento de la barrera
sanguínea del nervio óptico. Arch Soc Esp Oftalmol. 2016;91:170–176.
∗
Corresponding author.
E-mail address: raleman@dmor.ulpgc.es (R. Alemán).
2173-5794/© 2016 Sociedad Española de Oftalmología. Published by Elsevier España, S.L.U. All rights reserved.
 a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(4):170–176 171

Resultados: Los microvasos de los nervios ópticos de los animales diabéticos por efecto de la
estreptozotocina se caracterizaron por un incremento en el grosor de su pared, conservación
de los pericitos, incremento de la transcitosis en la célula endotelial y la presencia de una
población importante de macrófagos perivasculares. En general, las manifestaciones del
efecto de la hiperglucemia en el nervio óptico fueron más semejantes a las descritas para
la microcirculación cerebral que a las descritas para la retina.
© 2016 Sociedad Española de Oftalmología. Publicado por Elsevier España, S.L.U. Todos
los derechos reservados.

a single intraperitoneal injection of streptozotocin (65 mg/kg)

Introduction in citrate tamponade with a pH of 4.5. A group of control ani-
mals with the same characteristics was submitted to a single
Diabetic optic neuropathy is a clinical condition involving the
tamponade injection. The level of glycemia reached by dia-
optic pathway of diabetic patients. It can course asymptomat-
betic animals exceeded 350 mg/dl, whereas in control animals
ically, appearing only in neurophysiological studies, and it
said level was under 110 mg/dl. The diabetic animals were not
can also appear without the presence of diabetic retinopathy.
administered insulin for correcting hyperglycemia.
Although the etiopathogenic mechanism of said neuropathy
The glucose levels in plasma were assessed weekly by
is not well established, many studies indicate vascular origin.
means of an oxidase glucose reaction system (Menarine Glu-
Anterior optic neuropathy involves small choroidal circulation
cocard test, Glucocard International, Florence, Italy).
vessels that feed the anterior portion of the optic nerve before
After 6 and 12 weeks, 18 diabetic animals and 18 controls
reaching the lamina cribrosa, whereas posterior optic neurop-
were sacrificed after being anesthesized with an intraperi-
athy, which seems to be more frequent in diabetic patients,1
toneal injection of equithesin 3.3 ml/kg (9.5 mg pentobarbital
could mainly involve microvessels reaching the nerve from the
sodium, 76 ml ethanol, 42.5 mg chloral hydrate, 428 mg propyl-
pial circulation network.
ene glycol, 21 mg magnesium sulfate, and distilled water up to
Taking into account the above and even though the vas-
1 l).
cular origin of diabetic neuropathy in any organ is broadly
The circulation system was washed by means of cardiac
accepted,2 the structural changes of vessels in this disease
perfusion with physiological saline solution in order to sub-
are largely unknown. On the other hand, the demonstrated
sequently fix the animal. Two types of fixatives were utilized:
production of inflammatory cytokines due to the oxida-
2% glutaraldehyde solution in 0.1 M, pH 7.4 phosphate tam-
tive stress taking place in diabetes3 has determined the
ponade, and 4% paraformaldehyde solution, depending on
possible therapeutic usefulness of steroids and nonsteroid
whether the samples were allocated for ultrastructural study
anti-inflammatories for treating diabetic neuropathy.
or immunofluorescence techniques, respectively. The optic
In a previous article4 the authors demonstrated morpho-
nerves were obtained through craniotomy and preserved in
logical alterations in optic nerve fibers in diabetes-induced
the same fixative utilized for perfusion. The experiment was
rats by means of streptozotocin. Said alterations included
carried out following the guidelines of RD 1201/2005 dated
damage to, and the disappearance of, large size nerve
October 10 (Spain) on the protection of animals utilized for
fibers and the presence of myelin remains inside the nerve.
research and other scientific purposes.
However, the authors have not found studies relating the
The optic nerves fixated with glutaraldehyde were post-
alterations found in the optic nerve fibers with vascular
fixed in OsO4 1%, dehydrated in ethanol and included in agar
origin.
resin 100 by means of a routine technique. One-half micron
The objective of this paper was to study the morphologi-
sections were stained with toluidine blue in borax and ultra-
cal, structural, morphometric and functional changes of optic
fine 500 Å sections with uranyl acetate and lead citrate. The
nerve blood vessels and the blood barrier in an experimen-
nerves fixated in paraformaldehyde were included in paraffin
tal diabetes model, to determine whether modifications took
molds.
place in the microvessels of hyper-glycemic animals that could
explain the alterations found in optic nerve fibers. In the case
of finding said alterations, comparing said changes with those Morphometry
described in the microvessels of the retina and brain as a con-
sequence of hyperglycemia. The morphometric studies of optic nerve vessels were car-
ried out utilizing both types of sections. The analysis was
made over ultrafine section photographs with an overall mag-
Materials and methods nification of 7200×, obtained with a Philiphs 301 electronic
transmission microscope (Royal Philips Electronics, Amster-
Animals and treatment of samples dam, Holland).
The intraneural vessels were selected randomly, measuring
Hyperglycemia was induced in a group of two-month old basement membrane, endothelium area, vascular perimeter
Sprague Dawley rats with a weight of 263 ± 58.65 g by means of and gauge of each vessel.
172 a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(4):170–176
Twenty-four intraneural vessels of the diabetics group and
the control group were selected for study. Four measure-
ments of the thinnest basement membrane region of each
vessel were made, excluding basal lamina areas that were tan-
gentially cut or merged with pericytes. The measured areas
were separated by a linear distance of at least 1 cm.
In addition, 4 measurements of the endothelial cell area
and vascular gauge perimeter area were taken. The endothe-
lial cell area was calculated on the basis of surface unit. The
endothelial cell area was considered in relation to the vascular
gauge perimeter.
The vascular area was obtained from the gauge areas in
every case.
The analysis was performed utilizing the Media Cybernet-
ics Imagepro 1986 system (Media Cybernetic Inc, Rockville,
USA).
All intraneural vessels were counted in the sections stained
with toluidine blue, and the overall area of the complete sec-
tions of each nerve was measured.
The results are shown with mean values and typical devi-
ations obtained for each group. A comparative study was
performed applying the Mann–Whitney test when comparing
2 samples, and the Kruskal–Wallis nonparametric test when
comparing more than 2 samples.
Immunofluorescence
The presence of albumin was determined by means of indi-
rect immunofluorescence in optic nerve sections fixated with
paraformaldehyde and embedded in paraffin.
The nerve sections were incubated with the rabbit human
anti-albumina antibody (Sigma Laboratorios S.L., Fuenlabrada,
Madrid, Spain) at a dilution of 1:32 in a humid chamber
overnight. The negative control was performed with sections
cultured with normal rabbit serum. After incubation, the sec-
tions were rinsed with PBS 3 times and incubated with the
second anti-rabbit goat serum antibody marked with fluo-
rescein isothiocyanate (FITC) at a dilution of 1:100 at room
temperature during 30 min. Subsequently, the sections were
rinsed with TPBS and fixated with PBS glycerol.
The MHC class II cells were detected with direct
immunofluorescence. The sections were incubated with anti-
Ia (OX6) monoclonal antibody conjugated with FITC (Sigma
Laboratorios S.L., Fuenlabrada, Madrid, Spain) at a dilution of
1:50 during 30 min at 37 ◦ C, following the previous protocol.
The samples were analyzed with a Nikon labophot-2 fluores-
cence microscope (Nikon Instruments, Amsterdam, Holland).
Results
Morphological studies
Control group
Cross-sectioned vessels exhibited flattened endothelial cells
linked by means of tight links forming a continuous endothe-
lium. In general, endothelial cells are surrounded by pericyte
prolongations of considerable length. Pericytes are separated
from endothelial cells by a basement membrane, excepting
where there is contact between both cell types. Pericytes are
Fig. 1 – Transmission electronic microscope. Optic nerve
and vessel of a control animal: endothelial cells are joined
by tight links and surrounded by pericytes (pc), which
maintain contact with endothelial cells.
surrounded by their basal lamina and by prolongations of
astrocytes linked by intercellular unions. Vessel spaces are
narrow (Fig. 1). No albumin was observed around the vessel
walls (Fig. 2a).
Diabetic group
Occasionally, endothelial cells adopt irregular morphologies.
It is possible to observe narrow intercellular links and gap
junctions; in general, there is no lack of union between cells
(Fig. 3a). The cytoplasm exhibits some dense bodies, abundant
vesicles open toward the non-luminal surface of the endothe-
lium and fusion of vesicles forming intra-endothelial channels
(Fig. 3b and c).
The pericytes are preserved, no signs of the lesion in these
cells is observed and in general they do not lose contact with
the endothelial cell due to hyperglycemia (Fig. 4a and b). Albu-
min can be observed around the vascular wall (Fig. 2b).
In animals, 6 weeks after the streptozotocin injection,
perivascular cells are abundant. These cells are located in
the perivascular space between the basal lamina of the per-
icytes and the basal lamina of the astrocytic prolongations,
where well-developed organelles, vacuoles, dense bodies and
secondary lysosomes can be observed (Fig. 5). These cells are
different from pericytes and are OX6-positive. These cells were
less abundant in diabetic animals 12 weeks after the injection.
Morphometric studies
The data obtained by morphometric studies are shown in
Table 1.
No differences were found in the transversal area of nerves
or the number of intra-neural vessels of the optic nerves of
diabetic and control animals.
The endothelial cell area per surface unit was greater
in the vessels of optic nerves of diabetic animals 6 weeks after
the streptozotocin injection than in control animals. This area
 a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(4):170–176 173

control animals. Said thickening was larger in diabetic animals

vis-à-vis control animals at 6 and 12 weeks.
The patency area of vessels in diabetic animals did not
exhibit significant changes at 6 or 12 weeks post-injection in
comparison to control animals.

Discussion

The present study illustrates the characteristics of optic nerve

Fig. 2 – Detection of albumin in optic nerves: (A) control and
(B) diabetic, through immunofluorescence. Staining can be
observed around some vessels in diabetic animals nerves.
increased with the duration of hyperglycemia. No differences
were found in the endothelial cell area of control animals
between 6 and 12 weeks after the tamponade injection.
The basal lamina was thicker in optic nerve vessels at 12
weeks then at 6 weeks after the injection, both in diabetic and
endoneurial vessels of animals with streptozotocin-induced
diabetes. These vessels are characterized by increased
wall thickness vis-a-vis controls, conservation of pericytes,
increased transcytosis in endothelial cells and a considerable
population of perivascular macrophages.
Vascular wall thickening occurs due to a marked basal
lamina thickening. The secretion of connective tissue growth
factor5 seems to play an important role in this thickening, as
well as increased thickness of endothelial cells. These 2 fac-
tors are generally accepted as characteristic of diabetic patient
micro-vessels. A close correlation has been described between
said thickening and diabetic neuropathy severity.6 In hyper-
glycemic patients the thickening of the basal lamina occurs
even before the expression of the disease, and has been related
to the loss of axons in early stages of peripheral diabetic
neuropathy.7
The present study demonstrated that the increase of the
basal lamina thickness occurs also with the aging of both dia-
betic and control animals. This thickness is always greater in
hyperglycemic animals. In this regard, it could be concluded
that hyperglycemia produces premature aging of blood vessels
in the optic nerve. Endothelial thickening has been described
in diabetic vessels in different types of neuropathies.8,9 The
present study demonstrated that this alteration is also present
in the utilized experimental model. In contrast with the
changes occurring in the basal lamina, the endothelial thick-
ening was not related to the age of the animal but with
the evolution of hyperglycemia. Accordingly, both factors,
i.e., basal lamina thickening and endothelial area thickening,
has been considered to be measures indicating neuropathy

Fig. 3 – Transmission electronic microscope. Ultrastructural characteristics of endothelium in optic nerves of hyperglycemic
animals: (A) endothelial cell exhibits large development of intra-cytoplasmatic membranous organelles. (B) A pericytary
process and some collagen fibers located between the endothelium and the perivascular cell. (C) The union of
intra-endothelial vesicles creates a channel communicating the luminal and abluminal zone of the blood vessel (I). bm:
basement membrane; db: dense body in perivascular cell; ec: endothelial cell; I: intraendothelial channel; L: vascular
patency; pc: pericytary process; pvc: perivascular cell; ui: inter-endothelial union.
174 a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(4):170–176

Fig. 4 – Micro-photographies show perivascular cells in the vessels of diabetic animals: (A) perivascular cell cytoplasm
containing lysosomes and dense bodies. (B) Perivascular cell with cytoplasm displaced to an end of the cell. asc: astrocyte;
bm: basement membrane; col: collagen fibers; L: vascular patency; pc: pericyte; pvc: perivascular cell; pvs: perivascular
space.

severity.6,9 The thickening of the vascular wall could hin-
der the exchange of substances and accordingly (as is the
case in peripheral diabetic neuropathy6 ) could account for
the presence of damaged nerve fibers in the optic nerves of
rats with streptozotocin-induced diabetes as well as for the
neurophysiological alterations in the optic nerve, even before
the expression of the disease.
A number of factors have been related with the increase of
basement membrane thickness and diabetes. Ruptured and
injured pericytes is one of said factors. In contrast with the
Fig. 5 – Prolongation of a pericyte (*) that maintains contact
with an endothelial cell in a nerve capillary of a diabetic
animal after 12 weeks. The endothelial cell is thinner in the
contact zone. ast: astrocyte; pc: pericyte.
observations in peripheral diabetic neuropathy,7 no pericyte
degeneration was observed in the studied optic nerves. In
addition, pericytes remained in contact with endothelial cells
in the majority of vessels. On the other hand, the degeneration
of pericytes is one of the earliest histological changes encoun-
tered in diabetic retinopathy and has been involved in the
formation of microaneurysms and neovessels in the retina.
Váldez et al.10 pointed out that the loss of pericytes is sufficient
cause to trigger the degeneration of microvessels in the retina.
In this regard, the behavior of vessels in the optic nerve is more
similar to that of brain vessels, in which pericyte degeneration
as a consequence of diabetes11 has not been described, than to
the behavior of retina or peripheral neuropathy vessels on the
hyperglycemic conditions.7 In the retina, the degeneration of
pericytes occurs even prior to the development of proliferative
diabetic retinopathy.12
The present study observed traces of albumin around vas-
cular perimeters as well as the formation of intra-endothelial
channels due to the merger of plasmatic vesicles, which
confirms increased transcytosis in the endothelial cells of
hyperglycemic vessels. Transcytosis, which involves the trans-
port of plasmatic proteins to adjacent cells and tissue,13 also
occurs in hyperglycemic vessels in other locations.14 The
substances transported to the extra vascular space could con-
tribute to the increase of the basement membrane thickness.
In the perivascular space surrounding the basal lamina,
a considerable number of perivascular cells was observed
with phagocytic activity expressing class II molecules of the
major histocompatibility complex. Said molecules expressed
in macrophages but not in pericytes. Perivascular cells,
which have the function of eliminating lipids and substances
having high molecular weights, appear in spaces around
cerebral vessels and seem to be a normal component of brain
 a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(4):170–176
Table 1 – Morphometric data of endoneural optic nerve vessels of diabetic and control animals.
Control
6 weeks
Nerve area 0.27 ± 0.038
Number of vessels 42.6 ± 5.4
Endothelial cell thickness 0.27 ± 0.03 0.29
Basal lamina thickness 0.056 ± 0.01 0.076
Vascular patency area 19.7 ± 15.2 16.8
a
Significant difference with 6-week control group (p < 0.05).
b
Significant difference with 12-week control group and with 6-week diabetic group (p < 0.05).
c
Significant difference with 12-week control group (p < 0.05).
microvascularization derived from pial cells and reaching the
brain. The presence of these cells has been related to
the degradation of extracellular matrix components,
increased vascular permeability and the existence of myelin
remains. These could express MCH II class of molecules in
damaged brains of rats and humans. In addition, it has been
pointed out that these are related to immunological reactions
occurring in the blood–brain interface and in neurodegener-
ative diseases. The present study demonstrates an increase
in transcytosis and, in a previous study, degeneration of
myelinic fibers in streptozotocin-induced diabetes.4 Both fac-
tors could justify the increase in the number of perivascular
cells with phagocytic activity. These cells could account for
the elimination of waste, with the perivascular space being
considered as a drainage pathway of soluble and insoluble
nervous system materials.
The density of endoneurial blood vessels in the optic nerve
at 12 and 6 weeks after streptozotocin injection did not exhibit
significant variations in the study, in contrast with the obser-
vations reported by Thrainsdottir et al.7 in the peripheral
nerves in initial stages of hyperglycemic disease.
Mostly, the findings of the present study did not match the
structural modifications of microvessels observed in diabetic
retinopathy. However, they do match the findings described for
cerebral vascular lesions produced by hyperglycemia. In this
regard, it has been demonstrated that hyperglycemia produces
basal lamina thickening and increased patency in the brain
microvessels of rats15–18 and dogs.12
On the other hand, the results of the present study also
confirm the existence of vascular modifications related to the
damaged nerve fibers of observed in a previous study4 and
provide data about the role played by vascular damage in dia-
betic optic neuropathy in humans.
Conflict of interests
No conflict of interests has been declared by the authors.
Acknowledgment
The authors wish to acknowledge Ms. Francisca Pérez Afonso,
regrettably post-humously, for the excellent technical assis-
tance provided during the course of this study.
175
Diabetic
12 weeks 6 weeks 12 weeks
0.3 ± 0.036 0.25 ± 0.06 0.26 ± 0.05
42 ± 8.29 40 ± 6.7 38.6 ± 8.6
± 0.02 0.35 ± 0.03a 0.51 ± 0.06b
± 0.01a 0.094 ± 0.01a 0.102 ± 0.01c
± 7.9 17.4 ± 16.3 15.65 ± 11.8
references
1. Fernández DC, Pasquini LA, Dorfman D, Aldana Marcos HJ,
Rosentein RE. Early distal axonopathy of the visual pathway
in experimental diabetes. Am J Pathol. 2012;180:303–13.
2. Greene DA, Sima AA, Stevens MJ, Feldman EL, Lattimer SA.
Complications: neuropathy, pathogenetic considerations.
Diabetes Care. 1992;15:1902–25.
3. Sandireddy R, Yerra VG, Areti A, Komirishetty P, Kumar A.
Neuroinflammation and oxidative stress in diabetic
neuropathy: futuristic strategies based on these targets. Int J
Endocrinol. 2014. Article ID 674987.
4. Alemán-Flores R, Mompeó Corredera B, Ortega Santana F,
Limiñana Cañal JM, Castaño González I. Fibras del nervio
óptico y diabetes experimental. Arch Soc Esp Oftalmol.
2000;75:315–20.
5. Kuiper EJ, van Zijderveld R, Roestemberg P, Lyons KM,
Goldschmeding R, Klaassen I, et al. Connective tissue growth
factor is necessary for retinal capillary basal lamina
thickening in diabetic mice. J Histochem Cytochem.
2008;56:785–92.
6. Britland ST, Young RJ, Sharma AK, Clarke BF. Relationship of
endoneurial capillary abnormalities to type and severity of
diabetic polyneuropathy. Diabetes. 1990;39:909–13.
7. Thrainsdottir S, Malik RA, Dahlin LB, Wiksell P, Eriksson KF,
Rosén I, et al. Endoneurial capillary abnormalities presage
deterioration of glucose tolerance and accompany peripheral
neuropathy in man. Diabetes. 2003;52:2615–22.
8. Timperley WR, Boulton AJ, Davis-Jones GA, Jarrat JA, Ward JD.
Small vessel disease in progressive diabetic neuropathy
associated with good metabolic control. J Clin Pathol.
1985;38:1030–8.
9. Malik RA, Masson EA, Sharma AK, Lye RH, Ah-See AK,
Compton AM, et al. Hypoxic neuropathy: relevance to human
diabetic neuropathy. Diabetología. 1990;33:311–8.
10. Valdez CN, Arboleda-Velasquez JF, Amarnani DS, Kim LA,
D’Amore PA. Retinal microangiopathy in a mouse model of
inducible mural cell loss. Am J Pathol. 2014;184:2618–26.
11. Kern TS, Engerman RL. Capillary lesions develop in retina
rather than cerebral cortex in diabetes and experimental
galactosemia. Arch Ophthalmol. 1996;114:306–9.
12. Wong HC, Elts SM, Phillips JW, Williams KA. Differential
growth of brain and retinal bovine pericytes. Diabetologia.
1992;35:818–27.
13. Simionescu M, Gafencu A, Anthoe F. Transcytosis of plasma
macromolecules in endothelial cells: a cell biological survey.
Microsc Res Tech. 2002;57:269–88.
14. Mompeó B, Popov D, Sima A, Constantinescu E, Simionescu
M. Diabetes-induced structural changes of venous and
arterial endothelium and smooth muscle cells. J Submicrosc
Cytol Pathol. 1998;30:475–84.
176 a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(4):170–176

15. Kida S, Steart PV, Zhang E, Weller RO. Perivascular cells act as
scavengers in the cerebral perivascular spaces and remain
distinct from pericytes, microglia and macrophages. Acta
Neuropathol. 1993;85:646–52.
16. Frank RN, Dutta S, Frank SE. Cerebral cortical capillary
basement membrane thickening in galactosemic rats.
Diabetologia. 1987;30:739–44.
17. Huber JD, van Gilder RL, Houser KA. Streptozotocin-induced
diabetes progressively increases blood–brain-barrier
permeability in specific brain regions in rats. Am J Physiol
Heart Circ Physiol. 2006;291:H2660–8.
18. Prakash R, Johnson M, Fagan SC, Ergul A. Cerebral
neovascularization and remodeling patterns in 2 different
models of type 2 diabetes. PLOS ONE. 2013;8:e56264.