J. Cell Sci. 55, 199-210 (1982)
Printed m Great Britain © Company of Biologists Limited 1982
I go,
THE EUGLENA PARAFLAGELLAR ROD:
STRUCTURE, RELATIONSHIP TO OTHER
FLAGELLAR COMPONENTS AND PRELIMINARY
BIOCHEMICAL CHARACTERIZATION
JEREMY S. HYAMS
Department of Botany and Microbiology, University College London,
Goioer Street, London WCiE 6BT, U.K.
SUMMARY
Demembranated flagella from Euglena gracilis consisted of three distinct components: a 9 + 2
axoneme of microtubules, an extensive surface coating of long and short mastigonemes and a
lattice-like axial fibre known as the paraflagellar rod. Negatively stained preparations of
isolated paraflagellar rods showed the major structural unit to be a 22 nm filament oriented at
approximately 45° to the long axis of the rod in a 7-start left-handed helical arrangement.
Attachment of the paraflagellar rod to the flageliar axoneme was via a series of goblet-shaped
projections, which anchored the rod to a single outer doublet microtubule. Comparisons of
axonemes from E. gracilis bearing the paraflagellar rod with those of Chlamydomonas reinhardtii, in which it is absent, by polyacrylamide gel electrophoresis revealed the presence of
two prominent additional peptides with molecular weights of 80000 and 69000. Exposure
of Euglena axonemes to trypsin selectively solubilized the paraflagellar rod and also removed
both proteins, which were therefore tentatively identified as the major subunit proteins of the
rod and designated PFR 1 and 2, respectively. In addition to these details of the structure and
composition of the paraflagellar rod, the mode of attachment to the axoneme of both long and
short mastigonemes was also identified.
INTRODUCTION
In addition to the 9 + 2 axoneme of microtubules, many flagella possess various
types of axial accessory fibres. In several instances these reveal a paracrystalline
organization. This is particularly true of the mitochondrial derivatives, which form
a conspicuous component of the spermatozoa of certain insects (Baccetti et al. 1977).
Paracrystalline accessory fibres are also a feature of the flagella of some protozoa.
In euglenoid (Mignot, 1966; Leedale, 1967) and kinetoplastid (Fuge, 1969; Vickerman & Preston, 1976)flagellates,the diameter of the flagellum may be increased by
the presence of a lattice-like structure known as the paraflagellar rod. In Euglena
gracilis, the rod extends the length of the emergent flagellum. It assumes no consistent
orientation with respect to the central-pair microtubules, but appears to be the
attachment site for one of the two classes of mastigonemes that form an extensive
surface coating on these flagella (Bouck, Rogalski & Valaitis, 1978).
The function of the paraflagellar rod is unknown. It has been claimed that the
rod possesses ATPase activity and hence participates actively in flageliar movement
(Picclnni, Albergoni & Coppellotti, 1975; Bouck & Green, 1976). Others have sug-
200
J. S. Hyams
gested that it may be involved in the control of flagellar frequency and waveform
(Nichols & Rikmenspoel, 1980; Nichols, Jacklet & Rikmenspoel, 1980). A critical
assessment of these, and other, possibilities requires a detailed understanding of the
composition of the paraflagellar rod and its interaction with both the axoneme and
mastigonemes. This paper is addressed to both of these questions.
MATERIALS AND METHODS
Cell culture
Euglena gracilis (strain 1224/sz) and Chlamydomonas reinhardtii (strain 11/32!)) were both
obtained from the Cambridge Collection of Algae and Protozoa. Cultures of E. gracilis were
grown in a medium consisting of o-i % beef extract, 0-2 % yeast extract, 0-2 % Bacto tryptone
(all from Difco), o-i % sodium acetate and o-ooi % CaCl| (BDH) at 20 °C on a 14:10 light/
dark cycle. C. reinhardtii was grown on minimal medium (Sueoka, i960) under the same conditions of temperature and illumination but with continuous aeration.
Preparation of axonemes
Cultures (3 1) oi E. gracilis were grown to a density of ~ i o ' cells ml" 1 and harvested
by centrifugation at 2500 g for 5 min. Pellets were resuspended by swirling in a small volume
of growth medium and the cells further harvested by centrifugation for 5 min at ioooy in
50-ml glass conical tubes. Pellets were washed in HMDEK buffer, consisting of: 30 mMHEPES (A/-2-hydroxyethyl piperazine-AT'-2-ethanesulphonic acid) adjusted to pH 7-4 with
KOH, 5 mM-MgSO4, 1 mM-dithiothreitol, 0-5 mM-Na, EDTA, 25 mM-KCl at room temperature and resuspended in 00 ml of the same solution. The cell suspension was quickly
poured into 10 ml of ice-cold HMDEK containing 2-0 % Triton X-100 (Sigma), with rapid
stirring. This treatment resulted in > 95 % deflagellation. Cell bodies were removed by
centrifugation at 2000 g for 5 min and the supernatant containing the detached axonemes was
centrifuged further at 25000 g for 30 min at 4 °C. Axonemes from C. reinhardtii were prepared
using the same protocol, with the exception that the concentration of Triton X-100 was 0-5 %.
Electron microscopy
For thin sections, pelleted axonemes were fixed for 30 min in 2-s % glutaraldehyde in
HMDEK and washed three times in HMDEK. In certain instances the fixation solution
contained 8 % (w/v) tannic acid (Mallinckrodt). Following dehydration in an acetone
Fig. 1. Low-power electron micrograph of an axoneme negatively stained with uranyl
acetate and showing the extensive coating of long (/) and short (J) mastigonemes.
X5300.
Fig. 2. High-power view showing details of flagellar structure following staining
with gold thioglucose. The paraflagellar rod (pfr) is seen as a lattice-like structure
tightly associated with the microtubular axoneme (ax). Periodic projections can be
seen extending from the doublet microtuble to the right of the axoneme, i.e. diametrically opposite the paraflagellar rod. The dense covering of mastigonemes is evident,
x 43000.
Fig. 3. Transverse sections of axonemes showing the hollow nature of the paraflagellar rod. x 80000.
Fig. 4. As Fig. 3 but fixed with tannic acid, x 40000.
Fig. 5. Detail of mastigoneme attachment to the Euglena axoneme. Long mastigonemes
apparently arise in tufts from between the stalked projections that anchor the short
mastigonemes to the axoneme. x 60000.
Euglena paraflagellar rod
O
201
202
J. S. Hyams
series, axonemes were embedded in Araldite and sectioned on an LKB UMIV ultramicrotome.
Sections were counterstained with 2 % aqueous uranyl acetate and Reynolds'lead citrate and
examined in a Siemens Elmiskop 102 electron microscope. For negative staining, axonemes
were washed by centrifugation and resuspension in HMDEK to remove traces of detergent,
and placed on ionized carbon/Formvar grids prior to staining with either 2 % aqueous uranyl
acetate or 1 % gold thioglucose.
Sodium dodecyl sulphate (SDS)/polyaerylamide gel electrophoresis
Pellets of axonemes were resuspended directly in sample buffer (Stephens, 1975) and boiled
for 2 min. Electrophoresis was on 5 % polyacrylamide gels (6 mm x 8 cm) according to the
method of Stephens (1975). Gels were stained by the procedure of Fairbanks, Steck & Wallach
(1971) and molecular weights were calibrated with oligomera of bovine serum albumin (Payne,
1973) and other standard proteins. Gels were scanned using a Joyce-Loebl Chromoscan, and
the relative proportions of different bands were determined by cutting out and weighing the
areas under the appropriate peaks.
Trypsinization of axonemes
Axonemes in HMDEK were exposed to digestion by 20 ftg ml" 1 trypsin (Sigma) in either
of two ways: (1) axonemes immobilized on carbon/Formvar electron-microscope grids were
inverted onto a drop of trypsin solution on a Teflon block pre-equilibrated to 30 CC. After
varying periods of time the reaction was terminated by running several drops of the appropriate
negative stain over the grid; (2) for bulk digestion, suspensions of axonemes were incubated at
30 °C in 20 fig ml"1 trypsin. Changes in turbidity were monitored using a Pye Unicam SP 1800
8pectrophotometer. At 50 % of the initial turbidity the reaction was terminated by the addition
of a tenfold excess (w/w) of soybean trypsin inhibitor (Sigma). Digested axonemes were
harvested by centrifugation and prepared for electrophoresis as described above.
RESULTS
Ultrastructure of the Euglena axoneme
Negatively stained axonemes revealed an extensive coating of both long and short
mastigonemes (Fig. 1). The latter appeared to be associated with a series of stalked
projections 62 run in length and spaced at intervals of 120 nm along the side of the
axoneme diametrically opposite to the paraflagellar rod (Figs. 2, 5). By contrast, the
long mastigonemes appeared to attach directly to the surface of the microtubule,
arising between the stalked projections in tufts of five to six (Fig. 5). The paraflagellar
rod in these preparations appeared as a lattice-like structure, intimately associated
with the axonemal microtubules (Fig. 2). In transverse section, the rod revealed a
hollow profile with an outer diameter of 90 nm (Fig. 3). The hollow nature is seen
more convincingly in axonemes fixed in the presence of tannic acid (Fig. 4), although
in these preparations the dimensions of the rod were somewhat larger. Attachment
of the rod appeared to be to a single outer doublet, possibly to the A tubule (Fig. 3).
Fig. 6. Isolated paraflagellar rod. x 19000.
Figs. 7, 8. High magnification views of isolated paraflagellar rods showing the gobletlike projections extending from one surface (arrows), x 43000.
Fig. 9. Fragmented axoneme in which the paraflagellar rod has remained attached to a
single outer doublet microtubule. The periodic nature of the attachment between the
two is indicated by the arrow, x 45 000.
Figs. 10, 11. Isolated paraflagellar rod. The goblet-like projections can be seen particularly clearly. Fig. io, x 38000; Fig. 11, x 70000.
Euglena paraflagellar rod
203
204
J. S. Hyams
Ultrastructure of the paraflagellar rod
Examination of the fine structure of the paraflagellar rod was facilitated by the
finding that, in rare instances, the rod dissociated cleanly from the other axonemal
components (Figs. 6-n). Images obtained from such preparations were consistent
with the rod being composed of a series of coiled filaments forming a 7-start left-
Fig. 12. Diagram illustrating the three-dimensional structure of the paraflagellar
rod as deduced from electron micrographs of isolated preparations. The projections
to the left attach the paraflagellar rod to one of the nine outer doublet microtubules
of the axoneme.
handed helix with a pitch of approximately 450 and a periodicity of 54 nm. Details of
the helical symmetry were obtained directly from micrographs, such as that shown in
Fig. 7, where only one side of the cylindrical rod is visualized. By contrast, the
dimensions of the helical filament were seen most clearly in preparations in which the
rod had flattened, and front and back images were superimposed (Fig. 11). In this
case the filament measured 22 nm in diameter. Extending from the surface of the rod
was a series of goblet-shaped projections, which are seen in Figs. 7 and 8 but are
particularly clear in Figs. 10 and 11. In frayed axonemes in which the rod remained
attached to a microtubule doublet, the projections were seen to form the point of
attachment between rod and microtubule (Fig. 9). In the example shown in Fig. 8,
Euglena paraflagellar rod
205
0
the paraflagellar rod has twisted through 180 , reversing the normal orientation of
the projections. It should be stressed that the isolated paraflagellar rods shown in
Figs. 6-8, 10 and 11 were extremely rare. In the vast majority of cases where axonemes
has spontaneously frayed or been mechanically sheared by several passages through a
syringe, the paraflagellar rod remained firmly attached to one doublet microtubule, as
seen in Fig. 9. A diagramatic representation of the structure of the paraflagellar rod
and its microtubule linkages is shown in Fig. 12.
Identification of paraflagellar rod proteins
Prior to the direct analysis of purified paraflagellar rod preparations, the composition of this structure was approached using two indirect methods. In the first, the
B
dz
-~-
—.
*•••'•
pfr /_
pfr2-
.--.*
13
14
Fig. 13. SDS/polyacrylamidegels of axonemea from E.gracilis (A)and C.remhardtii (B),
showing dynein (d), tubulin (t) and the paraflagellar rod proteins pfr 1 and 2.
Fig. 14. SDS/polyacrylamide gels of axonemes from E. gracilis before (A) and after (B)
digestion with trypsin. Note the disappearance of pfr 1 and 2. The low molecular
weight band at the bottom of gel (B) is the soybean trypsin inhibitor used to terminate
the digestion.
J. S. Hyams
206
40
iDynein
30
20
T
o
•5. 10
8
-
6
Tubolin
I 4
Soybean trypsin inhibitor*?
0-5
10
Relative mobility
1-5
Fig. 15. Molecular weight calibration on SDS/polyacrylamide gels. (#) Oligomers
of bovine serum albumin. The dynein bands migrate in the region denoted by the
bracket.
SDS/polyacrylamide gel pattern of Euglena axonemes was compared to parallel
preparations of a related organism lacking the parafiagellar rod, namely, the green
alga, C. reinhardtii. In so doing it was assumed that the axonemes of both organisms
would reveal similar gel patterns and that, since Euglena mastigonemes are not
solubilized by SDS (Bouck et al. 1978; our unpublished observations), prominent
additional bands in the Euglena electrophoretogram could be ascribed tentatively to
the subunits of the parafiagellar rod. The results of such an experiment are shown
in Fig. 13. As expected, the two gels revealed many similarities, the major difference in
the Euglena gel being the presence of a pair of bands migrating behind the prominent
tubulin band (Fig. 13). The two species, which represent 16-9% of the total axonemal
protein, have been tentatively designated parafiagellar rod proteins (PFR 1 and 2).
The two are present in the proportion of 1:1-27. Co-electrophoresis against standard
proteins yielded molecular weights of 80000 and 69000, respectively (Fig. 15). A
summary of a number of molecular weight calibrations of axonemal proteins is
presented in Table 1.
A second approach to the molecular composition of the paraflagellar rod exploited
Euglena paraflagellar rod
207
Table 1. Molecular weight of flagellar proteins as determined by
gel electrophoresis
Protein
Dynein C
Dynein B
PFRi
PFRz
Tubulin
Molecular weight
(xio~')
337±u
3i3±6
8o±2
55 ±2
Number of determinations.
SDSjpolyacrylamide
n*
3
3
4
4
6
its sensitivity to digestion by the proteolytic enzyme trypsin. Suspensions of axonemes
incubated with trypsin showed a rapid decline in absorbance at 350 nm to approximately 50% of the initial value (Fig. 16). In the electron microscope, the decline in
absorbance was found to be correlated with the selective solubilization of the paraflagellar rod, the mastigonemes and axonemal microtubules being apparently little
affected (Figs. 17-19). Polyacrylamide gels of axonemes harvested before and after
treatment with trypsin are shown in Fig. 14. Although a number of minor bands
appear to be reduced in intensity, the major change associated with loss of the paraflagellar rod was the disappearance of PFR 1 and 2.
Fig. 16. Change in turbidity {A,«,) versus time for axonemes in HMDEK with (#)
and without (O) trypsin. Trypsin was added after 1 min.
J. S. Hyams
208
17
Figs. 17-19. Effect of trypsin digestion on axoneme structure.
Fig. 17. Control, x 3800.
Fig. 18. Axoneme incubated with trypsin. Note that the diameter of the axoneme
appears to be considerably reduced. Patches of short mastigonemes have started to
peel away (arrows), x 3800.
Fig. 19. High-magnification view of a region from Fig. 18. The paraflagellar rod has
been completely digested, x 46000.
DISCUSSION
Although extensively studied by thin-section electron microscopy (Mignot, 1966;
Fuge, 1969), the detailed structure of the paraflagellar rod of both euglenoid and
kinetoplastid flagellates as revealed by negative staining has received scant attention
(Bouck et al. 1978; de Souza & Souto-Padr6n, 1980). Application of this technique to
isolated paraflagellar rods from E. gracilis has shown the major structural unit to be a
22 nm diameter filament oriented obliquely to the long axis of the rod, generating
a 7-start left-handed helix. Although the regularity of the axial spacings argue for
a subunit construction, intermolecular boundaries within the filament have yet to be
resolved. The major periodicity of the paraflagellar rod revealed by this study was
54 nm. In his description of the paraflagellar rod of Trypanosoma brucei, Fuge (1969)
reported a major spacing, which was half this figure. Similar values were obtained
Euglena paraflagellar rod
209
from other trypanosomes (de Souza & Souto-Padr6n, 1980). This appears to be one of
a number of genuine differences in the organization of the paraflagellar rod in the
two groups of organisms. In E. gradlis the rod is hollow, whilst in T. brucei it contains
up to seven longitudinal filaments (Fuge, 1969). Further, the paraflagellar rod of
trypanosomatids maintains a strict relationship to theflagellaraxoneme, being attached
to doublets 4 to 7, whilst we, and also Bouck et al. (1978) have shown that in E. gradlis
the rod does not assume a consistent orientation with respect to the central
pair.
There are at least three possible explanations for the apparently random positioning
of the paraflagellar rod in relation to the axoneme. First, the rod could move systematically from doublet to doublet duringflagellarbeating. Second, during assembly,
the rod could become randomly but irreversibly associated with any of the nine
outer doublets. Third, the location of the rod could be fixed, but the orientation of the
central pair varied. At present it is not possible to distinguish between these alternatives. However, it has recently been reported that ciliary beating in Paramedum
is accompanied by rotation and twist of the central-pair microtubules (Omoto & Kung,
1980). Although central-pair rotation does not occur in cilia that generate a planar
waveform (Tamm & Tamm, 1981), it may be a feature of those organisms, like Euglena,
with a three-dimensional ciliary or flagellar beat.
A preliminary biochemical characterization of the Euglena paraflagellar rod has
identified two subunit proteins of 80000 and 69000 molecular weight, which were
designated PFR 1 and 2, respectively. Our study does not preclude the presence of
other minor components, nor does it reveal how these two major proteins are organized
into the helical structure of the rod. However, both of these questions may be resolved
when techniques for the isolation of pure preparations of paraflagellar rods become
available.
One previously unsuspected property of the rod revealed by this study was its
extreme sensitivity to trypsin. Digestion of the paraflagellar rod caused no apparent
change in the distribution of either the long or the short mastigonemes. This observation argues against the long mastigonemes inserting into the rod, as was suggested
previously (Bouck et al. 1978). Indeed, we have presented evidence that these mastigonemes attach directly to the outer doublet microtubules lying diametrically opposite
to the axoneme-rod junction. We have also identified the mode of attachment of the
short mastigonemes, which are anchored to stalked projections spaced at intervals of
120 nm, apparently along the same doublet as that from which the long mastigonemes
arise.
This study represents a first approach to the structure and function of the Euglena
paraflagellar rod. The conditions used for the isolation of axonemes in our experiments
were based on those shown to support the in vitro reactivation of Chlamydomonas
flagella (Allen & Borisy, 1974; Hyams & Borisy, 1975, 1978). We have found that
axonemes from Euglena will similarly reactivate upon the addition of ATP (unpublished observations). The fact that the paraflagellar rod may be selectively solubilized
may allow us to investigate its function by comparing the waveforms of axonemes in
the presence and absence of this organelle.
210
J. S. Hyams
I thank Adriana Luba for enthusiastic technical assistance and John Fagg for preparing
Fig. 12. This work was supported by Science Research Council grant GRA 97707 and a
Scientific Investigations grant from the Royal Society.
REFERENCES
C. & BORISY, G. G. (1974). Flagellar motility in Chlamydomonas: reactivation and
sliding in vitro. J. Cell Biol. 63, 5 a.
BACCETTI, B., DALLAI, R., PALLINI, V., ROSATI, F. & AFZELIUS, B. A. (1977). Protein of insect
sperm mitochrondrial crystals: Crystallomitin. J. Cell Biol. 73, 594-600.
BOUCK, G. B. & GREEN, P. M. (1976). Paracrystals and mastigonemes are directly attached to
the Euglena axonemal microtubules. J. Cell Biol. 70, 156 a.
BOUCK, G. B., ROGALSKI, A. & VALAITIS, A. (1978). Surface organisation and composition of
Euglena. II. Flagellar mastigonemes. J. Cell Biol. 77, 805-826.
DE SOUZA, W. & SOUTO-PADR6N, T. (1980). The paraxial structure of the nagellum of Trypanosomatidae. jf. Parasitol. 66, 229-235.
FAIRBANKS, G., STECK, T. L. & WALLACH, D. F. H. (1971). Electrophoretic analysis of the
major polypeptides of the human erythrocyte membrane. Biochemistry 10, 2606-2617.
FUGE, H. (1969). Electron microscopic studies on the intraflagellar structures of trypanosomes.
J. Protozool. 16, 460-466.
HYAMS, J. S. & BORISY, G. G. (1975). Flagellar coordination in Chlamydomonas reinhardtii;
isolation and reactivation of the flagellar apparatus. Science, N.Y. 189, 891-893.
HYAMS, J. S. & BORISY, G. G. (1978). Isolated flagellar apparatus of Chlamydomonas: characterization of forward swimming and alteration of waveform and reversal of motion by calcium
ions in vitro. J. Cell. Sci. 33, 235-253.
LEEDALE, G. F. (1967). EitgUnoidflagellates, New Jersey: Prentice-Hall.
MIGNOT, J. P. (1966). Structure et ultrastructure de quelques eugldnomonadines. Protistologica 2, 51-117.
1+
NICHOLS, K. M., JACKLET, A. & RIKMENSPOEL, R. (1980). Effects of Mg*+ and Ca on photoinduced Euglenaflagellarresponses..?. Cell Biol. 84, 355-363.
NICHOLS, K. M. & RIKMENSPOEL, R. (1980). Flagellar waveform reversal in Euglena. Expl Cell
Res. 129,377-381.
OMOTO, C. K. & KUNG, C. (1980). Rotation and twist of the central-pair microtubules in the
cilia of Paramecium. J. Cell Biol. 87, 33-46.
PAYNE, J. W. (1973). Polymerisation of proteins with glutaraldehyde. Soluble molecular-weight
markers. Biochtm.J. 135, 867-875.
PiCCiNNi, E., ALBERGONI, V. & COPPELLOTTI, O. (1975). ATPase activity inflagellafrom Euglena
gracilis. Localisation of the enzyme and effects of detergents. J. Protozool. aa, 331-335.
STEPHENS, R. E. (1975). High resolution preparative SDS-polyacrylamide gel electrophoresis:
fluorescent visualisation and electrophoretic elution-concentration of protein bands. Analyt.
Biochem. 65, 3&9~375SUEOKA, N. (i960). Mitotic replication of deoxyribonucleic acid in Chlamydomonas reinhardtii.
Proc. natn. Acad. Sci. U.S.A. 46, 83-91.
TAMM, S. L. & TAMM, S. (1981). Ciliary reversal without rotation of axonemal structure in
ctenophore comb plates. J. Cell Biol. 89, 495-509.
VICKERMAN, K. & PRESTON, T. M. (1976). Comparative cell biology of kinetoplastida flagellates.
In Biology of Kinetoplastida (ed. W. H. R. Lumsden & D. A. Evans), pp. 35-130. New York,
London: Academic Press.
ALLEN,
(Received 3 September 1981)