16

Supplementary Methods

Saccharomyces cerevisiae strains. YEF473, a gift from Professor Mark Longtine (Oklahoma State University), is the parent strain used in this paper. The result was reproduced in two other backgrounds, Hansen BY4743 (Open Biosystems), and L5976 (S1278b strain background, a gift from Professor Gerald Fink, Whitehead Institute).

L5976 (a/a ura3-52/ura3-52 trp1::hisG/trp1::hisG leu2::hisG/leu2::hisG); BY4743 (a/a his3-D1/his3-D1 leu2-D0/leu2-D0 ura3-D0/ura3-D0 met15-D0 lys2-D0); YEF473 (a/a ura3-52/ura3-52 lys2-801/lys2-801 leu2-D1/leu2-D1 his3-D200/his3-D200 trp1-D63/trp1-D63).

Molecular biology techniques and reagents. Diploid yeast expressing one C-terminally Gfp-tagged septin allele were created via the one step PCR tagging technique31, 32. The plasmid used for PCR amplification was pLP27, a pRS304 derivative containing the Gfp coding sequence (S65T, V163A; GenBank accession number AJ002683) between the XbaI and BamHI sites (a gift from Professor Rong Li laboratory, Harvard Medical School). Yeast transformations were performed as previously described33. The identity of the transformants was confirmed by sequencing (Harvard Biopolymers Facility; GfpR primer) the PCR amplified Gfp-tagged genomic septin allele. The Cdc3F (or Cdc12F) and GfpR primers were used for amplification. The primers’ sequences (5’ to 3’) are given in Supplementary Table 1.

For the in vitro analysis, the Gfp DNA was sub-cloned into a pET-28 vector (Novagen) between the BamHI and XhoI sites. The resulting vector was transformed into BL21 (DE3) bacteria. Gfp was isolated using standard procedures34 and further purified from large aggregates on a Superdex 200 gel filtration column (Amersham Biosciences). The monomeric species were used for anisotropy measurements.

Microscope considerations and calibration. The filter cube used for excitation/emission was the Endow GFP Bandpass (Chroma Technology Corp). During the image acquisition, the epifluorescence field diaphragm was minimized (diameter of approximately 25-30 mm), such as to reduce the amount of scattered light and improve the signal to background ratio. To correct for differences in transmission of the horizontally and vertically polarized light through the microscope optics the setup was calibrated using a fluorescein solution. Since fluorescein fluorescence lifetime is much longer than its rotational correlation constant, differences in IXX, IYY[i] are mainly due to differences in transmission. A correction factor of 0.9 was applicable for IYY. This transmission bias was a constant for our setup and was checked frequently during the course of the experiments. Differences in the light path resulting in misalignment were also accounted for. We used sub-resolution fluorescent beads (~170 nm, Component B of the PS-Speck Microscope Point Source Kit, Molecular Probes) and determined that the misalignment of the YY acquired image relative to the XX one was approximately one pixel in the X direction, and two pixels in the Y direction.

In vivo analysis considerations. The nocodazole arrested or partially arrested yeast are easier to image, quantify and interpret than the asynchronous population. Because the mother and daughter of the arrested yeast have similar sizes, the yeast population is packed uniformly with the mother-daughter axis in the stage plane. For the asynchronous culture, the packing can be non-uniform, many cells having the mother-daughter axis at some unknown angle to the microscope stage. The hourglass septin structures of middle to large budded yeast are quantitatively and qualitatively the same for an asynchronous and a nocodazole arrested population. The particular sets of data presented in this paper are from nocodazole arrested yeast, unless otherwise noted.

In vitro analysis. Yeast extract were prepared35 from the engineered strains and septin complexes were purified as previously described36. The purifed complexes were spun at 20,800 g for 5 minutes and diluted in 20 mM Tris pH 8, 1 M KCl. The Gfp anisotropy r=37 was determined using the Cary Eclipse fluorescence spectrophotometer (Varian). The correction factor G for our measurements (lex= 490 nm, 10 nm slit; lem= 510 nm, 10 nm slit) is 1.70. The final solute concentration was 20 mM Tris pH 8, 1 M KCl, 1% sucrose. The measured anisotropies are shown in Supplementary Table 2.

To form filaments and bundles, freshly eluted septin complexes were dialyzed against low salt buffer38. The bundles were tighter and more anisotropic when incubated with 2 mM MnCl2 for 30 minutes at 30o C or overnight at 4o C. Those bundles were used for the analysis presented in Fig. 3a and Supplementary Fig. 1c.

Supplementary Discussion

The mathematical model. A collection of Gfp dipoles attached to the septin scaffold describes a surface similar to an hourglass. We approximate the hourglass surface for the large budded cells to a hyperboloid of one sheet39. Parametric equations describing the average structure in large budded cells can be derived, given the following measurements. The average height of the structure (measured along mother-daughter axis) is 0.6 mm (standard deviation (SD)= 0.06 mm), while the average bud-neck diameter is 1.3 mm (SD= 0.1 mm). The slopes of the asymptotes to the surface generating hyperbola39 are ~/ 0.77. Such, the following parametric equation describes the surface:

Hyperboloid(u,v)=(0.65 cosh(v)cos(u), 0.65 cosh(v)sin(u), 0.5 sinh(v)), 0£u£2Pi and -0.56<v <0.56. The plot of this surface is shown bellow, as generated in Mathematica 5.2 (Wolfram Research, Inc.).

We engineered yeast strains to express fusion proteins between septins and Gfp such that the fluorophore is orientationally restricted relative to the septin protein. The resulted Gfp dipole ensembles displayed strongly polarized fluorescence. We developed a mathematical model to analyze the data based on the following observations and conclusions:

·  The order present at the Gfp dipole level suggests that the underlying septin structure is made of ordered septin subunits;

·  The little noise present in the population data (Fig. 2b, empty black squares) shows that the property we measure and probably the septin organization is the same from one yeast to another. We assume that each yeast lands on the gelatin pad randomly, since the yeast morphology presents no indication of potential bias. Thus the same result (Fig. 2b, empty black squares) suggests that the septin organization is invariant upon rotation around the mother-daughter axis (C symmetry axis). The uniform organization is also supported by the anisotropic pattern present in individual yeast imaged with the mother-daughter axis perpendicular to the microscope stage (Fig. 2a).

We define a right handed coordinate system, xyz, centered at each point on the hyperboloid surface ; x= dHyperboloid /du, y= dHyperboloid /dv, z= x Ä y[ii], evaluated at the particular (u, v) corresponding to each point, and normalized (Fig 1a). We consider a sub-resolution area around each point that is approximately flat and define f(q, f) as its static dipole distribution. The distribution function is normalized and described using the spherical coordinates q and f, as shown in the figure below.

The conformational motion of Gfp is slow relative to its fluorescence lifetime40 and in the designed septin-Gfp fusions its steady state anisotropy approaches 0.4 (Supplementary Table 2). Since the excitation and emission transition dipoles are approximately along the same direction within the Gfp coordinate frame40, we assume that the two dipoles are collinear during the Gfp fluorescence lifetime for our analysis.

The intensity of a particular area for the light path going through the excitation polarizer Ex (X or Y), and the emission polarizer Em (X or Y) is: IExEm=; DEx is the dipole projection on the Ex direction and DEm is the projection on the Em direction. K is a correction factor that incorporates lamp intensity, polarizers, microscope, and sample concentration differences. Since we are concerned with ratios (P) of the same sample field, we ignore K for the theoretical treatment.

There are a few considerations in quantifying and interpreting the data:

·  The dipole orientational distribution around the average;

·  The particular curved geometry of the structure as described above along with its size, close enough to the resolution limit of the microscope (~0.26 mm). Thus for any measurement, more than one unit area is averaged over a curved surface;

·  The excitation and emission depolarization caused by high numerical aperture objectives41. However, corrections for the emission depolarization can easily be incorporated into the model equations, while the excitation depolarization is negligible for our system;

·  Depolarization due to light scattering (more significant for the in vitro experiments).

The above mentioned reasons contribute to the observed depolarization (absolute value of P different from 1). The dipole distribution can be attributed to not perfect restriction of the Gfp fluorophore relative to the Cdc3 a-helix, to some flexibility of the a-helix relative to the septin complex, to imperfect alignment of the septin complexes, or to all of the above. Further consideration to this issue is given in the ‘Septin filament orientation within the hourglass’ section below.

Septin filament orientation within the hourglass. The in vitro-assembled septin bundles are three-dimensional arrangements of longitudinal filaments38. For the Cdc3 and Cdc12 strains described in the main text, the average dipole direction is at 90-degrees to the bundles (Fig. 3a, Supplementary Fig. 1c). However, since the filament assemblies (the bundles) probably have cylindrical symmetric, the average dipole direction relative to an individual septin filament may not be precisely at 90-degrees.

For the analysis that follows we use the Cdc12 strain since it is the most anisotropic and thus most useful in restricting the bud-neck filament orientation. Depending on the Gfp linkage flexibility, the average dipole orientation relative to an individual Cdc12 tagged filament (Supplementary Fig. 1c) could be between 63 to 117 degrees (the limits are reached for a rigid linkage). However, this interval is based on the maximum P measured in vitro and is likely much more narrow, since light scattering depolarization is significant for the in vitro experiments (data not shown). For the in vivo measurements, the hourglass structure P approaches unity (Supplementary Fig. 1a). This suggests that the Gfp dipoles have a tight distribution around their average, the x-axis. Thus the septin filaments should be oriented at 63 to 117 degrees to the x-axis in the xy plane, which is along the y-axis +/- 27 degrees (the maximum deviation from the y-axis is probably largely overestimated, as explained above).

The measurement. Due to the sub-resolution curvature of the septin scaffold, IXX and IYY reflect not only the dipole projections in the polarizers’ plane, but also the projection on the perpendicular axis. The most affected measurements are the ones in areas B and C. Since the axial (along the light path) resolution of the microscope is worse than the lateral one, areas B and C measurements average over a larger portion of the septin surface. Qualitatively, the (IXX- IYY) maximum is still reached when the yeast is oriented such that the average dipole projection is along the X-axis (calculation not shown). This justifies ignoring the quantitative aspect in the main text, which is concerned with finding the average dipole direction.

Supplementary Notes

31. Schneider, B. L., Seufert, W., Steiner, B., Yang, Q. H. & Futcher, A. B. Use of polymerase chain reaction epitope tagging for protein tagging in Saccharomyces cerevisiae. Yeast 11, 1265-74 (1995).

32. Tatchell, K. & Robinson, L. C. Use of green fluorescent protein in living yeast cells. Methods Enzymol 351, 661-83 (2002).

33. Gietz, R. D. & Woods, R. A. Transformation of yeast by lithium acetate/single-stranded carrier DNA/polyethylene glycol method. Methods Enzymol 350, 87-96 (2002).

34. Patterson, G. H., Knobel, S. M., Sharif, W. D., Kain, S. R. & Piston, D. W. Use of the green fluorescent protein and its mutants in quantitative fluorescence microscopy. Biophys J 73, 2782-90 (1997).

35. Cheng, S. C., Newman, A. N., Lin, R. J., McFarland, G. D. & Abelson, J. N. Preparation and fractionation of yeast splicing extract. Methods Enzymol 181, 89-96 (1990).

36. Vrabioiu, A. M., Gerber, S. A., Gygi, S. P., Field, C. M. & Mitchison, T. J. The majority of the Saccharomyces cerevisiae septin complexes do not exchange guanine nucleotides. J Biol Chem 279, 3111-8 (2004).

37. Lakowicz, J. R. Principles of Fluorescence Spectroscopy (Kluwer Academic/ Plenum Publishers, 1999).

38. Frazier, J. A. et al. Polymerization of purified yeast septins: evidence that organized filament arrays may not be required for septin function. J Cell Biol 143, 737-49 (1998).

39. Weisstein, E. W. CRC Concise Encyclopedia of Mathematics. (CRC Press, 1999).

40. Volkmer, A., Subramaniam, V., Birch, D. J. & Jovin, T. M. One- and two-photon excited fluorescence lifetimes and anisotropy decays of green fluorescent proteins. Biophys J 78, 1589-98 (2000).

41. Axelrod, D. Carbocyanine dye orientation in red cell membrane studied by microscopic fluorescence polarization. Biophys J 26, 557-73 (1979).

Supplementary Table 1

Primer sequences

The 5’ to 3’ primers’ sequences are shown. The primers ending in F are the forward primers, while those ending in R are the reverse primers. The names of the forward primers reflect the fusion proteins they generate. The notation Septin_aDb_Gfp is used to indicate that a amino-acids have been deleted from the septin C-terminus, and b amino-acids from Gfp N-terminus. The strain presented in the main text was obtained using the Cdc3_20D4_GfpF primer, while the strain presented in Supplementary Fig. 1 was obtained using the Cdc12_3D4_GfpF primer. Cdc3-GfpF and Cdc12-GfpF are the primers used for creating the respective full-length septin to full-length Gfp fusions, Cdc3-GlySer-GfpF and Cdc12-GlySer-GfpF are used for making the fusions with a two amino-acids linker between the full-length proteins.

Cdc3-GfpF / GTTAACCACTCCCCCGTCCCTACAAAGAAGAAGGGATTTTTACGTATGAGTAAAGGAGAAGAACT
Cdc12-GfpF / CTAGAAGAGCAGGTCAAAAGCTTGCAAGTAAAAAAATCCCATTTAAAAATGAGTAAAGGAGAAGAACT
Cdc3-GlySer-GfpF / GTTAACCACTCCCCCGTCCCTACAAAGAAGAAGGGATTTTTACGTGGATCCATGAGTAAAGGAGAAGAAC
Cdc12-GlySer-GfpF / CTAGAAGAGCAGGTCAAAAGCTTGCAAGTAAAAAAATCCCATTTAAAAGGATCCATGAGTAAAGGAGAAGAAC
Cdc12_3D4_GfpF / GTAAAAAAACTAGAAGAGCAGGTCAAAAGCTTGCAAGTAAAAAAATCCGAAGAACTTTTCACTGGAGTTG
Cdc3_20D4_GfpF / AATTAAAAGCTTTGGAGGACAAGAAAAAACAGCTAGAACTTTCAATAAATGAAGAACTTTTCACTGGAGTTG
Cdc3taggingR / ATAATATTTAATAGTGTATGTTTGAAATTTTTATATGTCTTTATTTCGCGTACAATCTTGATCCGGAGC
Cdc12taggingR / GATAGGCGTTGAAATTGACGAGACAAAGAGGAAGACATTAATTAATCACGTACAATCTTGATCCGGAGC
Cdc3F / TGAGGCAAAACTAGCCAAACTAGAAATTG
Cdc12F / GATTTCACTCTTCCAGCAATTGCACCA
GfpR / GATAATGGTCTGCTAGTTGAACGCTTC

Supplementary Table 2