305 / 412
parameters such as thickness, size, orientation, and location.
14
The ratio of collagen to elastic fibers also evolves, increasing
with age, and the predominance of the former explains the
rigidity and reduction of the vibratory mucosal wave. Such
alterations, combined with loss of elastin and hyaluronic
acid, further decrease the viscosity of vocal folds and nega-
tively affect voice emission in the elderly population.
1,4,5,12,15
Given the complexity and interdependence of vocal fold
microstructures, one can expect them to be susceptible to
organic changes with aging. Published studies focused more
on collagen types I and III, while the rest were little investi-
gated. Thus, the goal of our study was to analyze the con-
centration of collagen types I, III, IV, and V and of elastin
in human larynx senescence by immunohistochemistry.
Methods
The research project was approved by the Ethics Committee
for Research on Human Beings of Botucatu Medical
School/UNESP, Sa˜o Paulo, Brazil (reference 3861/2011).
Twenty larynges from elderly men (aged 60-90 years)
were included, as well as 10 from male adult controls (aged
30-50 years), collected at autopsy. The elderly group was
subdivided into 2 age subgroups: 60 to 75 years (n = 10)
and greater than 75 years (n = 10).
The cause of death was retrieved from the autopsy
records, and when necessary, additional information was
obtained from medical records and family. The exclusion
criteria were septicemia; prolonged intubation; systemic
infections; persistent dermatologic, autoimmune, or meta-
bolic diseases; smoking habit; and cervical trauma of any
kind or other conditions that might compromise the mucosa
of vocal folds and invalidate the immunohistochemical
analysis.
Fresh larynges were incised at their posterior portion and
examined macroscopically to exclude lesions. As a standard
procedure, the middle part of the right vocal fold was dis-
sected (
Figure 1
) and immediately embedded in 10% buf-
fered formaldehyde for 24 to 48 hours. Paraffin blocks were
prepared for histological slides (hematoxylin and eosin
[H&E] stain) and immunohistochemical reactions. The left
vocal fold was used in another study.
The following antibodies were used: collagen I (Col1A1,
1:2000 dilution; Dako, Carpinteria, California), collagen III (Col
3A1, dilution 1:1000; Dako), collagen IV (Col4A, 1:40 dilution;
Santa Cruz Biotechnology, Santa Cruz, California), collagen V
(Col 5A1, 1:100 dilution; Santa Cruz Biotechnology), and elas-
tin (1:200 dilution, Dako). The antigen preparation for immuno-
histochemical reactions of collagens I, III, and V and of elastin
was performed with 1% pepsin at pH 1.8 and incubated for 15
minutes at 60
"
C and then 30 minutes at 37
"
C. Blocking was
performed with 8% methanol and hydrogen peroxide, followed
by 3% Molico milk (Nestle´, Vevey, Switzerland). The second-
ary complex HRP EnVision (Dako) was added for 1 hour.
Diaminobenzidine was added for 5 minutes, and Harris hema-
toxylin was added for 20 seconds. For collagen IV antigenic
sites retrieval, we used a Pascal (Dako) pressure chamber for 3
minutes in the pretreatment solution Trilogy (Cell Marque,
Rocklin, California). Following the protocol recommended by
our laboratory, the positive control used for collagen I, III, and
V and elastin was a fragment of kidney and, for negative control
of these same antibodies, only buffered saline, replacing the pri-
mary antibody in a series of sections each sampling.
Protein Expression Analysis
Reading of histological slides was performed by 3 authors,
blindly and randomly. The slides were evaluated at random,
without knowledge of the groups and the age of patients,
using a light microscope from Zeiss (Axiostar plus, Carl
Zeiss do Brasil Ltda), at 40
3
.
The analyzed locations were the basement membrane,
the endothelium vessels, and the lamina propria (superficial
and deep layers). Because of the imprecise nature of the
boundaries surrounding the intermediate layer of the lamina
propria, we decided to divide the lamina propria into 2
layers (superficial and deep) to facilitate presentation of the
results. The size of the lamina propria was measured
between the basal membrane and muscle fibers. The thick-
ness of the lamina propria was divided into 2 portions sub-
jectively as the superficial layer, corresponding to the upper
portion and positioned just below the basement membrane,
and the deep layer, corresponding to the lower portion and
positioned just above the muscle fibers.
Measurement of antibodies involved quantification of the
level of brown staining of the lamina propria and basement
membrane structures. The area was set at 2
m
m, with per-
centage used to report results.
16,17
ImageJ software was
used in this analysis. We set the polygon as the selection
tool for the area of interest, choosing 3 fields at random,
and we set the Color Deconvolution plugins to HDAB and
Make Binary.
Statistical Analysis
To compare age groups, considering the response profile
assessed in 2 layers of the vocal fold (superficial and deep
layers), we used a parametric variance analysis combined
Figure 1.
Vocal fold showing the site of the fragment dissection.
Branco et al
80