ARTICLE IN PRESS
European Journal of Cell Biology 84 (2005) 181–188
www.elsevier.de/ejcb
The mechanism of granulocyte nuclear shape determination:
possible involvement of the centrosome
Ada L. Olins, Donald E. Olins
Department of Biology, Bowdoin College, 6500 College Station, Brunswick, ME 04101, USA
Abstract
Mature blood neutrophils (polymorphonuclear granulocytes) have characteristically complex nuclear shapes. The
human neutrophil nucleus generally possesses 3–4 lobes; the mouse neutrophil nucleus frequently resembles a twisted
toroid with a central hole. Myeloid tissue culture systems (e.g., human HL-60 and murine MPRO) can be induced to
differentiate in vitro towards neutrophils by addition of retinoic acid, exhibiting the characteristic nuclear shape
changes. Confocal immunostaining and thin-section transmission electron microscopic image data from differentiated
HL-60 and MPRO cells clearly demonstrate proximity of the centrosomal region (containing dynein, g-tubulin and CNap1) to regions of granulocytic nuclear indentations. In addition, the centrosomal region, flanked by the Golgi
apparatus, is shown to be present within the central hole of the toroidal mouse granulocyte nucleus. A role for the
centrosomal region and associated microtubules in molding granulocytic nuclear shape is suggested.
r 2005 Elsevier GmbH. All rights reserved.
Keywords: Neutrophil; Nuclear envelope; Lamins; Lamin B receptor; Heterochromatin; Centrosome; Microtubules
Introduction
The blood neutrophil represents the frontline of
defense against invading bacteria and fungi. Nuclei in
these cells are characteristically non-spherical: the
human neutrophil nucleus generally possesses 3–4 lobes
(Lee et al., 1999); the mouse neutrophil nucleus is
frequently ‘‘ring-shaped’’ with a central hole (Biermann
et al., 1999). Studies indicate that these unusually
shaped nuclei are more deformable than spheroid nuclei,
facilitating neutrophil passage through the blood vessel
endothelial lining and rapid migration through tissue
interstitial spaces (Lee et al., 1999; Park et al., 1977).
Abbreviations: ELCS, nuclear envelope-limited chromatin sheets;
LBR, lamin B receptor; MT, microtubules; NE, nuclear envelope; RA,
retinoic acid
Corresponding author. Fax: +207 725 3405.
E-mail address: dolins@bowdoin.edu (D.E. Olins).
0171-9335/$ - see front matter r 2005 Elsevier GmbH. All rights reserved.
doi:10.1016/j.ejcb.2004.12.021
Progenitor cells within bone marrow generally possess
spheroid nuclei, which undergo post-mitotic shape
changes and heterochromatin condensation (Bainton et
al., 1971). The mechanism of myeloid nuclear differentiation is largely not understood. Leukemic tissue
culture systems (e.g., HL-60 cells) can be differentiated
into granulocytic form in vitro by addition of retinoic
acid (RA), providing useful models for the process of
granulopoiesis. Several facts are known based upon
studies of HL-60 cells and human and murine genetic
disorders. These facts form the basis of current
speculation on the mechanism of myeloid nuclear
differentiation. (1) Undifferentiated and granulocytic
forms of HL-60 cells have a deficiency of lamins A/C
and B1, compared to monocytic forms (Olins et al.,
1998, 2001). (2) Lamin B receptor (LBR) is elevated in
both granulocytic and monocytic forms of HL-60 cells,
compared to the parent undifferentiated cells (Olins et
al., 2001, 2000). (3) Genetic deficiency of LBR in
ARTICLE IN PRESS
182
A.L. Olins, D.E. Olins / European Journal of Cell Biology 84 (2005) 181–188
humans and mice results in hypolobulation of blood
neutrophils (Hoffmann et al., 2002; Shultz et al., 2003).
(4) Granulocytic nuclear lobulation in HL-60 cells is
inhibited when the differentiating cells are exposed to
the microtubule (MT)-depolymerizing chemical, nocodazole; but not when exposed to the actin-depolymerizing chemical, cytochalasin D (Olins and Olins, 2004).
Based upon these observations, we have proposed a
model for neutrophil nuclear differentiation (Olins and
Olins, 2004), which postulates that: (1) the NE is
deformable, due to a paucity of lamins A/C and B1;
(2) the nuclear envelope is bound tightly to underlying
heterochromatin via elevated levels of LBR; (3) invaginations of the NE are mediated by intact cytoplasmic
MTs, possibly involving associated motor molecules.
Employing confocal immunostaining and thin-section
transmission electron microscopy, this study presents
image data documenting the close proximity of the
centrosomal region (with centrioles) to the major
nuclear invaginations of granulocytic HL-60 and to
the central hole of granulocytic MPRO cells.
Materials and methods
Cells
HL-60/S4 cells were obtained from Dr. A. Sartorelli
(Yale University, New Haven CT). When exposed to
1 mM RA, the vast majority of cells exhibit lobulated
nuclei by day 4 (Campbell et al., 1995; Leung et al.,
1992; Olins et al., 1998). The cells were cultivated and
differentiated as described earlier (Olins et al., 1998) in
RPMI 1640 made 5% with heat-inactivated fetal calf
serum. Cells were harvested for immunostaining and
electron microscopy after 4 days of exposure to RA.
MPRO (mouse promyelocytic) cells were purchased
Fig. 1. Confocal slices of immunostained HL-60/S4 cells. Columns: 0, undifferentiated (two left columns); RA (day 4), granulocytic
cells (two right columns). Antigens: blue, lamin B; red, C-Nap1; green, various antigens in the different rows. Rows: top, g-tubulin;
middle, dynein; bottom, Golgi (p58). Note the frequent merging of green g-tubulin and dynein with red C-Nap1 to generate yellow
centrosomal staining. ELCS (Olins et al., 1998) are recognizable as brightly blue stained patches of lamin B in the RA-treated cells.
ARTICLE IN PRESS
A.L. Olins, D.E. Olins / European Journal of Cell Biology 84 (2005) 181–188
from ATCC and cultivated as described by the
distributor (80% Iscove’s modified Dulbecco medium,
20% heat-inactivated horse serum, 10 ng/ml recombinant murine GM-CSF). Cells were made 10 mM RA and
harvested for immunostaining and electron microscopy
after 3 days exposure.
Antibodies and immunostaining
Goat anti-lamin B was obtained from Santa Cruz
Biotechnology Inc. (Santa Cruz, CA). Rabbit anticentrosomal protein C-Nap1 (Mayor et al., 2000) was a
gift from Dr. E. Nigg (M.P.I. Biochemistry, Martins-
183
ried). Mouse monoclonal antibodies against a-tubulin,
g-tubulin and Golgi (p58) were all purchased from
Sigma-Aldrich. Mouse monoclonal anti-cytoplasmic
dynein was purchased from Convance Research Products (Berkeley, CA). FITC-, Cy3-, Cy5-conjugated
donkey secondary antibodies were all purchased from
Jackson ImmunoResearch Laboratory, Inc. (West
Grove, PA). SlowFade was obtained from Molecular
Probes, Inc. (Eugene, OR).
Immunostaining experiments employed a previously
described procedure (Olins et al., 2000), with some
modifications: (1) Microscope slides were soaked overnight in 1/1 ethanol/ether and freshly coated with poly
Fig. 2. Transmission electron microscopy of RA-treated HL-60/S4 cells. These two examples of granulocytic cells reveal the Golgi
membranes surrounding the centriolar region, and in close proximity to the indentations of lobulated nuclei.
ARTICLE IN PRESS
184
L-lysine
A.L. Olins, D.E. Olins / European Journal of Cell Biology 84 (2005) 181–188
(MW 150–300,000; Sigma-Aldrich), just before centrifugation of the cells. (2) Slides were fixed in
methanol ( 20 1C, 10 min) followed by three washes in
PBS (5 min each). The cells were not excessively
flattened by this procedure. (3) No coverslip was used
during antibody incubations, to minimize loss of cells.
(4) Prior to the application of primary antibodies, slides
were incubated with 5% normal donkey serum (Jackson
ImmunoResearch Laboratory) in PBS for 15–30 min at
37 1C in a moist chamber. Confocal images were
collected on a Zeiss 510 Meta.
For Wright-Giemsa staining, cells were cytospun onto
ethanol-cleaned microscope slides, fixed in room temperature methanol for 15 min, air-dried and stained as
described earlier (Olins et al., 1998). For analysis of the
percentage of cells in the various nuclear morphology
categories, approximately 150 cells were observed and
classified in each experiment.
Electron microscopy
Cells were centrifuged and the supernatant medium
removed. The cell pellet was resuspended in 2.5%
glutaraldehyde in 0.05 M cacodylate (pH 7.2), 0.05 M
KCl, 2.5 mM MgCl2 for 2 h at room temperature. This
was followed by three washes in 0.05 M cacodylate (pH
7.2), 0.05 M KCl, 2.5 mM MgCl2. The cells were postfixed in 2% OsO4 in H2O for 2 h and washed with H2O.
The fixed pellet was suspended in 2% low melting
agarose, put on ice and cut into small blocks. Dehydration through 30%, 50% and 70% ethanol was followed
by overnight staining with 0.7% uranyl acetate in 70%
ethanol. The next day, dehydration was continued
through 80%, 90% and 95% ethanol on ice, followed
sequentially with 100% ethanol and propylene oxide at
room temperature. Epon prepared at 60 1C (Glauert,
1991) was used for the infiltration and embedding.
Fig. 3. Confocal slices of immunostained MPRO cells. Columns: 0, undifferentiated (two left columns); RA (day 3), granulocytic
cells (two right columns). Antigens: blue, lamin B; red, C-Nap1 (top row); green, various antigens in the different rows. Rows: top,
a-tubulin; middle, dynein; bottom, Golgi (p58). In the top row, note the merging of green a-tubulin with red C-Nap1 to generate
yellow centrosomal staining.
ARTICLE IN PRESS
A.L. Olins, D.E. Olins / European Journal of Cell Biology 84 (2005) 181–188
Blocks were cured at 60 1C for 12 to 24 h. Thin sections
were cut and stained with 2% uranyl acetate in
methanol and Reynold’s lead citrate. Images were
collected on a Philips 400 (German Cancer Research
Center, Heidelberg) and a Zeiss 10C (Marine Biological
Laboratory, Woods Hole, MA).
Results
HL-60/S4 differentiation
Nuclear shape changes have been quantified on
Wright-Giemsa stained, cytospun preparations of undifferentiated and RA treated (day 4) HL-60/S4 cells
(Olins et al., 2000). Undifferentiated cells exhibited the
following nuclear shapes: 83% ovoid, 15% indented
and 2% lobulated. By contrast, RA-treated cells
exhibited the following: 5% ovoid, 60% indented
and 35% lobulated. Cells from these two states were
cytospun onto polylysine-coated slides, fixed in metha-
185
nol without air-drying and immunostained with a
variety of antibodies directed against nuclear, centrosomal and Golgi components. Fig. 1 presents confocal
slices with merged colors from selected undifferentiated
and granulocytic HL-60/S4 cells. The images clearly
demonstrate the co-localization of g-tubulin, dynein and
C-Nap1, as well as close association of the Golgi antigen
(p58) with the centrosomal region (Colanzi et al., 2003).
In undifferentiated cells the centrosomal region is
juxtanuclear, often in a depression of the NE. In RAtreated granulocytic forms of HL-60/S4 cells, the
centrosomal and Golgi region are frequently between,
or adjacent to, deep nuclear invaginations. Binucleated
cells are also observed in the population of RA-treated
HL-60/S4 cells, with the centrosomal region usually
situated close to both nuclei.
Thin section transmission electron microscopy
(Fig. 2) of embedded and stained RA-treated HL-60/S4
cells reveals the close proximity of the centrosomal
region and the pericentriolar Golgi membranes to the
space between nuclear lobes (left, low magnification
images). At higher magnification (right images), Golgi
Fig. 4. Transmission electron microscopy of RA-treated MPRO cells. This gallery of low-magnification images illustrates the
apparent ultrastructural diversity of the granulocytic nuclear form. Panels a and b appear to be sections through primarily ringshaped nuclei. Panel c appears to have a lobulated nucleus; panels d–f look more like indented nuclear forms. Short examples of
ELCS can be observed in (b and c). Golgi/centrosomal regions can just be discerned at this low magnification in (a, c and f).
ARTICLE IN PRESS
186
A.L. Olins, D.E. Olins / European Journal of Cell Biology 84 (2005) 181–188
membranes can be observed surrounding the centrosomal region. The EM images are in good agreement with
the immunostained images of Fig. 1.
MPRO differentiation
Nuclear shape changes were quantified on WrightGiemsa stained, cytospun preparations of undifferentiated and RA-treated (day 3) MPRO cells. The
spectrum of nuclear shapes was more diverse than with
HL-60/S4 cells. Undifferentiated cells revealed the
following nuclear shapes: 64% ovoid, 18% indented
and 5% ring-shaped. About 13% of the cells exhibited
multiple nuclei, which were usually ovoid. RA-treated
cells exhibited the following nuclear forms: 7% ovoid,
38% indented, 45% ring-shaped and 8% lobulated. Multinucleated cells with ovoid nuclei were 2%.
Confocal immunofluorescent images of stained undifferentiated and RA-treated (day 3) MPRO cells are
presented in Fig. 3. As with HL-60/S4 cells, dynein and
C-Nap1 appear to co-localize in close proximity to the
NE. Ring-shaped nuclei, observed either in the undifferentiated or RA-treated cells, present dynein, C-Nap1
Fig. 5. Transmission electron microscopy of RA-treated MPRO cells. These two examples of granulocytic cells reveal the Golgi
membranes surrounding the centriolar region. Panels a and b may be illustrating an indented nucleus or an oblique section through
a ring-shaped nucleus. Panels c and d could be a cross-section of a ring or different nuclear lobes.
ARTICLE IN PRESS
A.L. Olins, D.E. Olins / European Journal of Cell Biology 84 (2005) 181–188
and Golgi (p58) within the ring hole. MTs (a-tubulin)
can be observed radiating from the centrosomal region
within the ring hole.
Thin section transmission electron microscopy of
embedded and stained RA-treated MPRO cells are
shown in Figs. 4 and 5. The images reveal the close
proximity of the centrosomal region and the pericentriolar Golgi membranes to the spaces within the ring
hole or nuclear indentations. These EM images are in
good agreement with the immunostained images of Fig.
3. Fig. 4 displays a low magnification montage of
nuclear shapes observed by random thin sectioning. Fig.
5 presents two examples where the centrioles can be
readily observed surrounded by Golgi membranes.
Random thin sections alone do not permit a precise
definition of nuclear shape. Although 3-D confocal
imaging facilitates a qualitative definition of shape,
serial section reconstruction or EM tomography would
be required to determine the exact granulocytic nuclear
shape, and the exact location of the Golgi and
centrioles.
Discussion
In vertebrate animals the major blood granulocyte
(‘‘neutrophil’’, in mammals) possesses distinctly nonspherical nuclei. Generally, this is regarded as an adaptation to permit these cells to migrate rapidly through
endothelial walls and interstitial spaces (Lee et al., 1999;
Park et al., 1977). Presumably these non-spherical nuclei
are more deformable than typical spherical nuclei. The
actual neutrophil nuclear shape appears to vary among
different species (e.g., humans exhibit lobulation; mouse,
twisted toroids or rings). Our present microscopic data
suggests considerable variation in granulocyte nuclear
shape within a particular species. Indeed, ring-shaped
nuclei have been reported in human blood smears from
infectious mononucleosis (Peichev, 1986), myelodysplastic
syndrome (Langenhuijsen, 1984; Stamen, 1985), Chagas’
disease (Cabral, 1987), multiple myeloma (Kanoh, 1991),
as well as normal individuals (Cabral and Robert, 1989).
Furthermore, examination of murine blood smears (data
not shown) reveals some neutrophil nuclei that look more
lobulated than toroidal. In addition, ring-shaped nuclei are
not confined to granulocytes in the mouse. They have also
been described in mouse bone marrow monocytic and
myeloid precursor cells (Biermann et al., 1999).
We have previously presented a working model of the
mechanism of nuclear differentiation during granulopoiesis (Olins and Olins, 2004). Many features of this
working model remain to be critically tested. The model
contains the following assumptions, based upon studies
of HL-60 cells and human and murine genetic disorders:
(1) a flexible NE (due to the paucity of lamins A/C and
187
B1) is ‘‘tacked down’’ to the underlying heterochromatin
(enhanced by elevated LBR); (2) the NE undergoes
invaginations in the vicinity of the centrosome due to
motor (e.g., dynein) attachment to the NE and movement along MTs towards the centriolar region; (3) new
NE membrane materials are added via lateral diffusion
from the ER (Holmer and Worman, 2001), resulting in
net membrane growth; (4) constraints on nuclear shape
by actin and spectrin-like proteins are weak, due to the
paucity of lamins A/C and NUANCE and the
cytoplasmic localization of emerin; (5) constraints on
nuclear shape by vimentin-envelope interactions may
play a role, but remains to be demonstrated in the
granulocyte system.
The goal of the present study is to document the
proximity of the centrosomal region and the Golgi
apparatus to the major invagination/nuclear hole of the
granulocyte nucleus. Previous publications contain electron micrographs also illustrating this issue: granulopoiesis in the rat (Tang and Clermont, 1989); granulopoiesis
in the mouse (Biermann et al., 1999). It is clear that this
documentation does not prove the working model.
Centrosomes embedded into the nuclear invagination/
hole may be an effect, rather than a cause, of the nuclear
shape change. ‘‘Self-centering’’ of the centrosome appears
to be a consequence of the ability of MTs to organize a
radial array around the cell center (Burakov et al., 2003).
Furthermore, it should be pointed out that in many cell
systems, mitotic chromosomes form ‘‘rosettes’’ around
spindle MTs (Allison and Nestor, 1999; Gerlich and
Ellenberg, 2003; Nagele et al., 1995). This observation
provokes a possible modification of our working model.
The timing of post-mitotic NE reformation (Burke and
Ellenberg, 2002; Gruenbaum et al., 2003), relative to
mitotic chromosome decondensation, may be different in
mouse, compared to human granulopoiesis. NE reformation while mitotic chromosomes possess rosette arrangement could result in ring-shaped nuclei. However,
delayed NE reformation, while the chromosomes are
fusing and decondensing, might yield a more spheroidshaped nucleus, requiring later participation of MTs and
the centrosomal region in the formation of the mature
lobulated granulocyte nucleus.
Acknowledgements
This work was supported by an NIH grant (R15HL075808) and departmental funds from Bowdoin
College. Some of the electron microscopy was carried
out, while the authors were Visiting Scientists in the
laboratories of P. Lichter and H. Herrmann (German
Cancer Research Center, Heidelberg). We wish to express
our appreciation for their generosity. This publication is
dedicated with great admiration and affection to our
friend for many years, Prof. Werner W. Franke.
ARTICLE IN PRESS
188
A.L. Olins, D.E. Olins / European Journal of Cell Biology 84 (2005) 181–188
References
Allison, D.C., Nestor, A.L., 1999. Evidence for a relatively
random array of human chromosomes on the mitotic ring.
J. Cell Biol. 145, 1–14.
Bainton, D.F., Ullyot, J.L., Farquhar, M.G., 1971. The
development of neutrophilic polymorphonuclear leukocytes
in human bone marrow. J. Exp. Med. 134, 907–934.
Biermann, H., Pietz, B., Dreier, R., Schmid, K.W., Sorg, C.,
Sunderkotter, C., 1999. Murine leukocytes with ringshaped nuclei include granulocytes, monocytes, and their
precursors. J. Leukoc. Biol. 65, 217–231.
Burakov, A., Nadezhdina, E., Slepchenko, B., Rodionov, V.,
2003. Centrosome positioning in interphase cells. J. Cell
Biol. 162, 963–969.
Burke, B., Ellenberg, J., 2002. Remodelling the walls of the
nucleus. Nat. Rev. Mol. Cell Biol. 3, 487–497.
Cabral, H.R., 1987. Neutrophils with ring-shaped nuclei in
Chagas’ disease. Br. J. Haematol. 67, 118–119.
Cabral, H.R., Robert, G.B., 1989. Ring-shaped nuclei in
human neutrophilic leukocytes of healthy individuals:
evidence of their occurrence and characteristics. Am. J.
Hematol. 30, 259–260.
Campbell, M.S., Lovell, M.A., Gorbsky, G.J., 1995. Stability
of nuclear segments in human neutrophils and evidence
against a role for microfilaments or microtubules in their
genesis during differentiation of HL60 myelocytes. J.
Leukoc. Biol. 58, 659–666.
Colanzi, A., Suetterlin, C., Malhotra, V., 2003. Cell-cyclespecific Golgi fragmentation: how and why? Curr. Opin.
Cell Biol. 15, 462–467.
Gerlich, D., Ellenberg, J., 2003. Dynamics of chromosome
positioning during the cell cycle. Curr. Opin. Cell Biol. 15,
664–671.
Glauert, A.M., 1991. Epoxy resins: an update on their
selection and use. Microsc. Anal. Sept., 13–19.
Gruenbaum, Y., Goldman, R.D., Meyuhas, R., Mills, E.,
Margalit, A., Fridkin, A., Dayani, Y., Prokocimer, M.,
Enosh, A., 2003. The nuclear lamina and its functions in the
nucleus. Int. Rev. Cytol. 226, 1–62.
Hoffmann, K., Dreger, C.K., Olins, A.L., Olins, D.E., Shultz,
L.D., Lucke, B., Karl, H., Kaps, R., Muller, D., Vaya, A.,
Aznar, J., Ware, R.E., Sotelo Cruz, N., Lindner, T.-H.,
Herrmann, H., Reis, A., Sperling, K., 2002. Mutations in
the gene encoding the lamin B receptor produce an altered
nuclear morphology in granulocytes (Pelger–Huet anomaly). Nat. Genet. 31, 410–414.
Holmer, L., Worman, H.J., 2001. Inner nuclear membrane
proteins: functions and targeting. Cell. Mol. Life Sci. 58,
1741–1747.
Kanoh, T., 1991. Ring neutrophils in plasma cell dyscrasia.
Arch. Pathol. Lab. Med. 115, 178–180.
Langenhuijsen, M.M., 1984. Neutrophils with ring-shaped
nuclei in myeloproliferative disease. Br. J. Haematol. 58,
227–230.
Lee, G.R., Foerster, J., Lukens, J., Paraskevas, F., Greer, J.P.,
Rogers, G.M., 1999. Wintrobe’s Clinical Hematology,
tenth ed. Williams & Wilkins, Baltimore.
Leung, M.-F., Sokoloski, J.A., Sartorelli, A.C., 1992. Changes
in microtubules, microtubule-associated proteins and intermediate filaments during the differentiation of HL-60
leukemia cells. Cancer Res. 52, 949–954.
Mayor, T., Stierhof, Y.D., Tanaka, K., Fry, A.M., Nigg, E.A.,
2000. The centrosomal protein C-Nap1 is required for cell
cycle-regulated centrosome adhesion. J. Cell Biol. 151,
837–846.
Nagele, R., Freeman, T., McMorrow, L., Lee, H.Y., 1995.
Precise spatial positioning of chromosomes during prometaphase: evidence for chromosomal order. Science 270,
1831–1835.
Olins, A.L., Olins, D.E., 2004. Cytoskeletal influences on
nuclear shape in granulocytic HL-60 cells. BMC Cell Biol.
5, 30.
Olins, A.L., Buendia, B., Herrmann, H., Lichter, P., Olins,
D.E., 1998. Retinoic acid induction of nuclear envelopelimited chromatin sheets in HL-60. Exp. Cell Res. 245,
91–104.
Olins, A.L., Herrmann, H., Lichter, P., Olins, D.E.,
2000. Retinoic acid differentiation of HL-60 cells
promotes cytoskeletal polarization. Exp. Cell Res. 254,
130–142.
Olins, A.L., Herrmann, H., Lichter, P., Kratzmeier, M.,
Doenecke, D., Olins, D.E., 2001. Nuclear envelope and
chromatin compositional differences comparing undifferentiated and retinoic acid- and phorbol ester-treated HL-60
cells. Exp. Cell Res. 268, 115–127.
Park, B.H., Dolen, J., Snyder, B., 1977. Defective
chemotactic migration of polymorphonuclear leukocytes
in Pelger–Huet anomaly. Proc. Soc. Exp. Biol. Med. 155,
51–54.
Peichev, M., 1986. Ring cells in infectious mononucleosis. Br.
J. Haematol. 62, 397–398.
Shultz, L.D., Lyons, B.L., Burzenski, L.M., Gott, B., Samuels,
R., Schweitzer, P.A., Dreger, C., Herrmann, H.,
Kalscheuer, V., Olins, A.L., Olins, D.E., Sperling, K.,
Hoffmann, K., 2003. Mutations at the mouse ichthyosis
locus are within the lamin B receptor gene: a single gene
model for human Pelger–Huet anomaly. Hum. Mol. Genet.
12, 61–69.
Stamen, P., 1985. Neutrophils with ring-shaped nuclei in
myeloproliferative disease. Br. J. Haematol. 59, 559.
Tang, X.M., Clermont, Y., 1989. Granule formation and
polarity of the Golgi apparatus in neutrophil granulocytes
of the rat. Anat. Rec. 223, 128–138.