A regulation is shown for size and number of the elements of complex ciliary structures forming the oral apparatus (OA) of a ciliate Paraurostyla weissei. Morphometric investigations were performed on oral ciliature of normal and size-reduced cells. Those constituents of the OA which exist as single structures, such as the inner and outer preoral membranelles, are not eliminated. Both shorten and the outer membranelle becomes narrower. Within the adornal zone of membranelles in size-reduced cells some frontal and ventral membranelles become eliminated, whereas the respective ratio of these types remains size invariant. In each individual adoral zone of membranelles there are membranelles of different length specificially located along the ventral part. Membranelles from small cells are significantly smaller than those of normal cells.

The number of kinetosomes is reduced in all four rows constructing an adoral membranelle. The analysis showed that regardless of cell size, the number of kinetosomes in the two inner rows of a membranelle is linearly and proportionately related. Regulation of the size of all components of the oral ciliature in P. weissei occurs at the time when the primordia of oral ciliature are formed.

The results are discussed in relation to recent ideas about pattern formation and size dependent regulation of the number and size of pattern elements.

As was recognized very early by Morgan (1901) and emphasized recently by Wolpert (1969) the question of whether the size or position of a part is proportional to the dimensions of the whole is theoretically veiy important. The precision of the size invariance of proportionality in patterns and the range of sizes across which the system can regulate have seldom been quantified.

The question has been studied carefully only on a limited number of animals. In Hydra the regulation of the internal body pattern understood as a composition of different types of cells (ectodermal, endodermal, nervous, etc.) comprises the number of pattern elements (Bode et al. 1973; Bode, Flick & Bode, 1977). Similarly the number of different cell types in the fruiting body of Dictyostelium varies proportionately with the size (Stenhouse & Williams, 1977). In Drosophila the bristle pattern was shown to be a function of cell number and cell size (Held, 1979). Interestingly in higher organisms the number of somites in early vertebrate embryos stays constant over a range of sizes, whereas the number of cells forming each somite varies (Cooke, 1975b; Flint, Ede, Wilby & Proctor, 1978; Pearson & Elsdale, 1979; Elsdale & Pearson, 1979).

These data have been generally interpreted as not fitting the general theory whereby patterns are specified by monotonic gradients of positional information as originally proposed by Wolpert (1969). Some such concept is nevertheless necessary to explain the global regulation of pattern elements in relation to overall size.

Ciliated protozoans being single-celled organisms offer special advantages for studying pattern formation and regulation in the absence of cellular partitions. They possess highly ordered arrays of cilia and ciliary aggregates with their fibrillar derivatives, known as the cortical pattern of the ciliature.

The object of our study is a hypotriche ciliate Paraurostyla weissei (Fig. 1) in which the ventral ciliature is composed of multikinetosomal aggregates: oral membranelles and cirri. Changes in the number of cortical elements take place only through replacing the whole set with a new one. In previous studies (Jerka-Dziadosz, 1976, 1977) it was found that the number of serially repeating structures such as left marginal cirri and adoral membranelles is variable and regulated proportionally to cell length. However, the effect of miniaturization of cell dimensions on the size and composition of elements of the complex pattern has not been studied in detail (Jerka-Dziadosz & Golinska, 1977). Constant size of oral membranelles in Stentor with concomitant reduction in their number was reported by Tartar (1941, 1967) whereas in starved Stylonychia cells Dembowska (1938) observed diminution of dimension of the individual cirri.

Fig. 1.

The relationship between the width (w) and the cell length (J) for 42 of P. weissei of different size. The line fitted by linear regression analysis is given by the equation: w = 5·2429 + 0·33401.

Fig. 1.

The relationship between the width (w) and the cell length (J) for 42 of P. weissei of different size. The line fitted by linear regression analysis is given by the equation: w = 5·2429 + 0·33401.

In this paper we present the results of analysis of pattern alterations of complex ciliary structures forming the oral apparatus in Paraurostyla weissei caused by reduction of cell dimensions. The ciliature is composed of frontal and ventral adoral membranelles and two preoral membranelles (previously called’undulating membrane’). The details of the ultrastructure of these parts of the oral apparatus were presented in a previous publication (Bakowska & JerkaDziadosz, 1978). Morphometric and statistical analysis was performed on changes of the cell form and ciliary pattern. The analysis of the mode of regulation of oral ciliature in P. weissei showed that adjustment of cortical pattern to cell dimensions comprises proportional changes not only in the number of the elements but also in their constitution - i.e. in the number and distribution of kinetosomes within elements.

The results are discussed in relation to the recent ideas about pattern formation and size dependent regulation of the number and size of pattern elements.

Cells used in these investigations were a clone of Paraurostyla weissei, line Z-6 isolated from single exconjugant after total conjugation (Jerka-Dziadosz & Janus, 1975). The culture methods were as previously described (Bakowska & Jerka-Dziadosz, 1978). Observations were made on cells from growing populations, on cells miniaturized by starvation in sterile Pringsheim solution where they passed through repetitive physiological reorganisations (Dembowska, 1938), or on cells miniaturized by repeated cutting of normal cells into several nucleated pieces which regenerated into small complete urostylas.

The counts of adoral membranelles were performed on preparations stained with protargol after Jerka-Dziadosz & Frankel (1969) under light microscope (Ortholux Leitz Wetzlar). Kinetosomes in membranelles were scored on ultrathin sections examined under an electron microscope JEM 100B. For electron microscopy, the ciliates were prepared as described previously (Bakowska & Jerka-Dziadosz, 1978). In some experiments cells were stained for 15-30 min in 0-1 % solution of tannic acid before dehydration. This procedure gives better contrast to microtubular structures.

As a measure of ciliary pattern alterations the following values were utilized:

(1) for the changes of the cell form - cell width to cell length ratio,

(2) for the changes of the adoral zone structure - the ratio of the number of frontal to ventral membranelles,

(3) for changes of architecture of membranelles - number and distribution of kinetosomes in rows in membranelles. Length of a membranelle is understood as the number of kinetosomes in its longest row, while the width of a membranelle is defined by the number of kinetosomal rows.

Statistical analysis of the results consisted of regression analysis and analysis of variance. Moreover, two tests for proportionality were used: (1) the 95% confidence interval of the F-intercept was computed. If it included the value X = 0, the relation was consistent with the conclusion of size invariance with 5 % probability of error, (2) Regression of the ratio X/Y on X was computed and tested whether it was significantly different from zero. If not, it was concluded that the ratio X/Y was the same at all values of X. Thus, the relation was size invariant. These are routine techniques and descriptionscan be found in Socal & Rohlf (1969) and Finn (1974). Preparation of the statistical programme and computations were performed by the computing Centre of the Polish Academy of Sciences.

I. Analysis of changes of cell dimensions caused by miniaturization

After several transections or continuous deprivation of food, Paraurostyla cells become progressively smaller. A question arises whether the cell dimensions change in such a way that the proportions between the cell width and length remain similar to the proportions characteristic of cells of normal size. To answer this question 41 individual cells of different size were measured. Some of these individual cells were fixed 3-5 to 4 h after division, some were size-reduced by operation, (i.e. they were fixed after termination of the regeneration processes about 5-6 h after the operation) and the remainder were starved for a few days. The results of the above measurements are plotted in Fig. 1. Statistical analysis showed that the cell width and length are strongly correlated (r = 0-843, P = 0·01) and this relation is of linear character. Both tests for proportionality showed that the ratio of the cell width vs. cell length is size invariant. It follows therefore that the miniaturized cells become shorter and proportionally slender.

The question then arises whether the relation between the analysed dimensions of the cell, and between cell size and the number of ciliary structures, is similar in naturally small (excystant) and experimentally size-reduced (operated or starved) cells. The results of measurements are presented in Table 1. The comparisons of linear regression lines calculated for the number of adoral membranelles against the cell length and regression of the cell width against the length revealed that the corresponding pairs of coefficients can be replaced by a common regression. From this comparison it follows that when the cell dimensions are similar, the regularities within the oral membranelles are the same regardless of the origin of the cell and the cause of size reduction.

Table 1.

Comparison of size-reduced and small excystant cells 0/Paraurostyla weissei

Comparison of size-reduced and small excystant cells 0/Paraurostyla weissei
Comparison of size-reduced and small excystant cells 0/Paraurostyla weissei

II. Analysis of changes in the zone of adoral membranelles

Size-reduced cells of P. weissei show a substantial reduction in the number of adoral membranelles. Statistical analysis showed strong correlation (r = 0·838, P = 0·01) between the total number of adoral membranelles and cell length (Jerka-Dziadosz, 1976). However, the zone of adoral membranelles is not structurally uniform (Bkowska & Jerka-Dziadosz, 1978). The frontal and ventral membranelles differ not only in structure but also in their location with respect to the antero-posterior axis of the cell. Frontal membranelles (‘collar’), (Figs. 2 and 3) are arranged in the form of an arc surrounding the anterior margin of the cell. They are composed of four rows of ciliated basal bodies (Fig. 4), with the shortest row possessing four to five kinetosomes, and the remaining three possessing 12-18 kinetosomes each. The cilia in the shortest row are much shorter than the others. The distances between frontal membranelles are larger than those between ventral membranelles.

Fig. 2.

The ventral surface of P. weissei. OA, oral apparatus; AZM, adoral zone of membranelles, hair-like appendages represent the ciliature of the ventral surface. Protargol staining, x 600, bar represents 10 μm.

Fig. 2.

The ventral surface of P. weissei. OA, oral apparatus; AZM, adoral zone of membranelles, hair-like appendages represent the ciliature of the ventral surface. Protargol staining, x 600, bar represents 10 μm.

Fig. 3.

The anterior part of the ventral surface of P. weissei. The AZM is divided in two parts: frontal and ventral. The line indicates the border between them. fAZM, frontal part of AZM; vAZM, ventral part of AZM; OPM, outer preoral mem brandie; IPM, inner preoral membranelle. Protargol staining, x 1800, bar represents 10 μm.

Fig. 3.

The anterior part of the ventral surface of P. weissei. The AZM is divided in two parts: frontal and ventral. The line indicates the border between them. fAZM, frontal part of AZM; vAZM, ventral part of AZM; OPM, outer preoral mem brandie; IPM, inner preoral membranelle. Protargol staining, x 1800, bar represents 10 μm.

Fig. 4.

Section through four frontal membranelles of a size-reduced cell. Numbers 1-4 enumerate kinetosomal rows, x 10000, bar represents 1 μm.

Fig. 4.

Section through four frontal membranelles of a size-reduced cell. Numbers 1-4 enumerate kinetosomal rows, x 10000, bar represents 1 μm.

Fig. 5.

Section through part of OPM and IPM of a normal cell. Numbers 1-5 enumerate kinetosomal rows of outer preoral membranelle. x 10 400, bar represents 1 /mi.

Fig. 5.

Section through part of OPM and IPM of a normal cell. Numbers 1-5 enumerate kinetosomal rows of outer preoral membranelle. x 10 400, bar represents 1 /mi.

The ventral membranelles are arranged (Fig. 7) parallel to each other and are oriented perpendicularly to the long axis of a cell. The membranelles are tightly packed, with the shortest anterior-most row of cilia in most cases possessing three kinetosomes. The border between frontal and ventral membranelles is located on the left anterior margin of the body (Fig. 4).

Fig. 6.

Section through OPM and IPM of a miniature cell. Numbers 1 and 2 enumerate the kinetosomal rows of the outer preoral membranelle. x 10000, bar represents 1 μm.

Fig. 6.

Section through OPM and IPM of a miniature cell. Numbers 1 and 2 enumerate the kinetosomal rows of the outer preoral membranelle. x 10000, bar represents 1 μm.

Fig. 7.

Section through the ventral part of AZM of a miniaturized cell. 1-4 enumerate kinetosomal rows, x 4600, bar represents 1 μm.

Fig. 7.

Section through the ventral part of AZM of a miniaturized cell. 1-4 enumerate kinetosomal rows, x 4600, bar represents 1 μm.

In older to gain some information about the relation between the two parts of AZM, the number of frontal and ventral membranelles was counted separately in 106 cells of different size. Statistical analysis showed that the numbers of both kinds of membranelles are strongly correlated and they are linearly related to each other. The regression line is superimposed in the scatter diagram in Fig. 8. Both tests for proportionality showed that the ratio between the number of frontal to ventral membranelles is size invariant. Hence, it can be concluded that when the cell dimensions decrease, the number of frontal and ventral membranelles correlates in such a way that the proportions between these parts of AZM remain constant regardless of the cell size.

Fig. 8.

The relationship between the numbers of frontal (f) and ventral (v) membranelles of normal and size-reduced cells of P. weissei. The line fitted by the linear regression analysis is given by the equation: fm = 0·012 + 0·3820vm.

Fig. 8.

The relationship between the numbers of frontal (f) and ventral (v) membranelles of normal and size-reduced cells of P. weissei. The line fitted by the linear regression analysis is given by the equation: fm = 0·012 + 0·3820vm.

III. Analysis of size variability of ventral membranelles of normal and size-reduced cells

It is known from the previous study (Jerka-Dziadosz, 1976), that the total number of adoral membranelles in size-reduced cells is smaller than in the normal ones. The question arises whether the number of kinetosomes in membranelles is also reduced in miniature cells. In order to answer this question two groups of cells were investigated. All cells designated to be of normal size were from a growing population. Therefore they had an average number of adoral membranelles of about 64 (Jerka-Dziadosz, 1976). The group of sizereduced cells was also relatively uniform in respect to the number of adoral membranelles. They possessed from 36 to 44 adoral membranelles. For further analysis only the ventral part of AZM was taken. The zone of the ventral membranelles, regardless of the cell size, was arbitrarily divided into six regions (Fig. 9). Membranelles found on sectioned material were classified to the particular region on the basis of occurrence of characteristic combinations of structures situated in vicinity of each region of the AZM. In no case was it possible to count kinetosomes in all membranelles from one cell. In our material we found 67 complete membranelles from 12 cells of normal size and 98 membranelles from 13 size-reduced cells. Therefore, the analysed membranelles of many cells form a ‘random population’.

Fig. 9.

Schematic representation of the ventral part of the membranellar band divided into six sectors numbered from anterior to posterior.

Fig. 9.

Schematic representation of the ventral part of the membranellar band divided into six sectors numbered from anterior to posterior.

Each membranelle regardless of the cell size is composed of four rows of kinetosomes (Figs. 7 and 11). The two posterior rows (R-l and R-2) are the longest and usually possess an equal number of kinetosomes. The third row (R-3) is always shorter than the two posterior ones. The most anterior (R-4) is the shortest and usually consists of only three kinetosomes. The number of kinetosomes in the three posterior rows is variable and depends upon the position of a membranelle within a zone (Table 2). The longest membranelles are located in the central regions, i.e. Ill and IV, the shortest are located at the extremities, i.e. in regions I and VI. Membranelles composed of identical number of kinetosomes similarly arranged into four rows occur in large as well as in small cells. However, such membranelles are differently located in vAZM of compared groups. One of the smallest membranelles from a large cell is located at the anterior end of region I and possesses 50 kinetosomes distributed: 17, 17, 13 and 3 in four successive rows, whereas in a small cell the membranelle corresponding in number and distribution of kinetosomes is located in the region III. Comparison of ranges of kinetosome number in the inner two rows (R-2 and R-3) of ventral membranelles in normal and size-reduced cells indicated that membranelles from small cells appear proportionately smaller.

Table 2.

Range of distribution of kinetosome number in ventral adoral membranelles of different AZM region

Range of distribution of kinetosome number in ventral adoral membranelles of different AZM region
Range of distribution of kinetosome number in ventral adoral membranelles of different AZM region

In order to test whether this relation really occurs, two way analysis of variance was used. Differences in number of kinetosomes in second rows of membranelles between normal and size reduced cells, difference between region I vs. Ill and IV, and possible interactions were simultaneously tested. Both the differences of membraneliar length (the number of kinetosomes in one of the two first rows) between regions of AZM and between normal and size-reduced cells proved to be highly significant (i-value of 24-9 and 12-9 respectively, D.F. = 17). The interaction, i.e. the degree to which the differences in one factor (regions) are not the same across the other (normal vs. size-reduced cells), was small (2-00) and not statistically significant with a 5 % probability of error.

Two conclusions arise from the above observations: first-that in each individual AZM band there are membranelles of different length specifically located along the ventral part, and second - that membranelles from small cells are significantly smaller than those from normal cells.

IV. Analysis of the number of kinetosomes and their distribution in normal and size-reduced cells

In order to define the cytogeometrical rules governing the architecture of ventral membranelles, the distribution of kinetosomes in the two inner rows (R-2 and R-3) was further analysed. From Table 2 it appears that membranelles show a rather constant relationship in the distribution of kinetosomes within the two rows. This possibility was tested statistically by regression analysis of kinetosome number in inner rows of membranelles from normal and sizereduced cells (Table 3). The analysis showed that the number of kinetosomes in the third row (R-3) is linearly related to the number of kinetosomes in the second row (R-2), in membranelles from both normal and size-reduced cells.

Table 3.

The statistical analysis of the internal distribution of kinetosomes within membranelles of normal and size-reduced cells of Paraurostyla weissei

The statistical analysis of the internal distribution of kinetosomes within membranelles of normal and size-reduced cells of Paraurostyla weissei
The statistical analysis of the internal distribution of kinetosomes within membranelles of normal and size-reduced cells of Paraurostyla weissei

The tests for proportionality showed that the relationship between the number of kinetosomes in the two inner rows is proportional. The comparison of regression equations showed that the two groups do not differ significantly. Relation between the number of kinetosomes in R-3 and R-2 in membranelles of normal and size-reduced cells can be described by a common linear regression (Table 3, Fig. 10). This relation is proportional.

Fig. 10.

The relationship between the number of kinetosomes in the two inner rows of kinetosomes of ventral membranelles. The common line for membranelles of normal and miniature cells fitted by linear regression analysis is given by the equation r3 = 1·0731+0·7434 r2.

Fig. 10.

The relationship between the number of kinetosomes in the two inner rows of kinetosomes of ventral membranelles. The common line for membranelles of normal and miniature cells fitted by linear regression analysis is given by the equation r3 = 1·0731+0·7434 r2.

Fig. 11.

Membranelles of the same relative position in the ventral band (sector posterior IV) from a normal cell (A) and a miniature cell (B), x 11500, bar represents 1 μm.

Fig. 11.

Membranelles of the same relative position in the ventral band (sector posterior IV) from a normal cell (A) and a miniature cell (B), x 11500, bar represents 1 μm.

In two exceptionally small cells the fourth (R-4) row in several membranelles has only two kinetosomes. The total number of ventral membranelles was lower than 20 in both cells. The cell defined as S had 14 ventral membranelles, so an extrapolation based on the results of Fig. 8 gives a total number of all membranelles equal to 19. This indicates a three-fold diminution in the size of the oral apparatus of this cell as compared to an average normal cell. The fourth row in all but the first anterior ventral membranelles was two-kinetosomal. The number of kinetosomes in rows R-l, R-2 and R-3 was also drastically reduced. The cell defined as U was slightly larger than the cell 5. It possessed about 19 ventral membranelles, implying a total of about 27 adoral membranelles. The R-4 in ventral membranelles of this cell is either three-or twokinetosomal with no specific pattern of arrangement of these membranelles within the zone. The above facts indicate that when cells of P. weissei undergo a decrease in the size of the membranelle band of more than 2-5 fold, the number of kinetosomes in the anterior rows of each membranelle is also reduced from three to two. At sizes near threshold, the transition between the two patterns is gradual rather than sudden.

V. Relation between the length and position of membranelles within the AZM and the internal distribution of kinetosomes in membranelles

From the analysis described in previous sections it follows that some basic rules of patterning of the ventral band of membranelles should be the same in normal and size-reduced cells, i.e., they should be size-independent.

There are two possibilities regarding the architectural regularities in the structure of membranelles of normal and size-reduced cells:

(1) the distribution of kinetosomes in the two inner rows is the same within the membranelles of the same length (the membranellar length is understood as the number of kinetosomes in R-l) regardless of the cell size,

(2) membranelles, regardless of their length with similar position in the adoral zone show similarrelationshipindistribution ofkinetosom es in the two inner rows.

In order to check the first hypothesis, a statistical analysis was performed on membranelles of size-reduced and normal cells possessing from 24 to 33 kinetosomes in each of the first two rows. Since the membranelles in VI region are different from the other, membranelles having less than 24 kinetosomes in the first row were excluded in this analysis. The results are presented in Table 3. The comparison of the regression lines for membranelles of the same length of normal and size-reduced cells showed statistically significant differences between them. The value of the regression coefficient is higher for membranelles from miniature cells.

From this analysis it follows that membranelles of the same length but of different locations within the AZM of cells of different size have different internal patterns.

In order to test the second hypothesis, which postulates that membranelles of different length but of similar locations within adoral zones show a similar relationship between numbers of kinetosomes in thé two inner rows, membranelles from region III and IV (from normal and size-reduced cells) were chosen for statistical analysis and comparison.

The analysis of membranelles possessing 29-41 kinetosomes in the second row, derived from normal cells, showed linear relation (Table 3). This relation is not proportional.

In small cells, the membranelles located centrally in the adoral zone have 25 to 33 kinetosomes in the second row (Table 2). As in the previous cases numbers of kinetosomes in the second and third row are linearly related, but the relation is not proportional (Table 3).

Similar analysis of kinetosomal distribution performed separately for cells U and S and cell U alone showed also that there exists a linear relation. However, the proportions between the number of kinetosomes in third vs second row do not depend upon the membraneliar length (Table 3).

The value of regression coefficients calculated for analysed groups of centrally located membranelles are similar. In order to find out whether the differences between them are significant or not the test for equality of slopes of several regression lines was used. This showed that all of regression coefficients b can be replaced bya common coefficient which is equal to 0-98794. In other words, when the number of kinetosomes in the second row (and the first as well) of centrally located membranelles decreases by one kinetosome then the third row shortens also by one kinetosome regardless of membranellar length. This indicates that the relationship in kinetosomal distribution between the inner row of these membranelles is relatively constant. It follows therefore that the position of a membranelle has a decisive effect on the value of the regression coefficient regardless of the length of that membranelle.

Summarizing the results obtained from comparison of the data concerning membranelles from normal and size-reduced cells, it can be stated that membranelles of the same length vary with respect to their inner pattern according to their position within AZMs during regulation of the size of oral apparatus in relation to overall cell dimensions. In other words, at least for central AZM locations, membranelles of similar position in the AZM are identical with respect to their inner pattern regardless of their length.

Another aspect of the proportions of membranelles concerns their total architecture, i.e the ratio between the number of kinetosomal rows in a membranelle and the length of the rows of kinetosomes. Since each membranelle is always composed of four rows of kinetosomes and the length of kinetosomal rows can vary, the shape of membranelles depends on membraneliar size. This means that membranelles from small cells corresponding in position with membranelles from normal cells are relatively shorter and thicker (Fig. 11).

The comparison of characteristics of membranelles from normal and sizereduced cells reveals regularities common for both groups of data (Table 2):

(1) The length range of membranelles of a given AZM is the larger the more membranelles it contains. Therefore, this range depends on the cell size.

(2) The value of the regression coefficients b close to 1 is a constant feature of kinetosomal distribution between the inner rows of centrally located membranelles in all kinds of cells.

(3) The smaller the total number of membranelles, the shorter are the membranelles located centrally.

(4) The number of ventral membranelles is highly correlated with the number of frontal membranelles and the total number of the adoral membranelles is in turn correlated with the length of a cell (Jerka-Dziadosz, 1976). It follows from the above that the number of ventral membranelles is also correlated with the cell length. This allows one to conclude that the distribution of kinetosomes in ventral membranelles depends on the overall dimension of the cell.

VI. Analysis of ultrastructural variability in preoral membranelles of normal and size-reduced cells

The right side of the oral opening in P. weissei is bounded by two longitudinal ciliary aggregates called collectively the ‘membrana undulans’ UM, (Figs. 2, 3). These are two separate structures : inner and outer ciliary rows (de Puytorac, Grain & Rodrigues de Santa Rosa, 1976; BAkowska & Jerka-Dziadosz, 1978).

The inner preoral membranelle (1PM) formed by a single row of ciliated kinetosomes originates on the ventral side of the oral opening at the same level as the most posterior AZM membranelles and runs anteriorly reaching a level of the border between II and III region of the ventral part of the AZM (Fig. 3).

The outer preoral (OPM) membranelle is formed by several longitudinal rows of alternatively distributed ciliated kinetosomes (Fig. 5) and is slightly shorter than the inner membranelle. The length and spacing of both membranelles is correspondingly smaller in size-reduced cells. Kinetosomes in both membranelles were not counted because it was impossible to obtain sections through all cilia even on serial sections.

If one accepts that the distance between kinetosomes of the UM’s are the same in normal and size-reduced cells and the length of the whole membranelle is correspondingly smaller in miniature cells it could be concluded that the number of kinetosomes in the inner preoral membranelle is proportionally coordinated with the size of the oral apparatus - that is with the size of the cell.

This conclusion - concerning the length of the membranelle - is true also for the outer preoral membranelles. In addition to this the outer membranelle is regulated also in its width. In normal cells the outer membranelle is usually formed by 4-5 longitudinal rows of kinetosomes (Fig. 5). The two rows on the AZM side are the longest and are usually equally long. The subsequent rows to the right of the preceeding ones do not reach both ends of the membranelle. This means that a membranelle is widest in its middle part.

In cells reduced in size by starvation or transection, outer preoral membranelles possessing only three or two rows of kinetosomes in the widest part were observed (Fig. 6). However, we have never observed an outer membranelle having only one row of kinetosomes. In our material, therefore, the maximal number of kinetosomal rows in normal cells is 6-7 and the minimal number of them in size-reduced cells is 2.

Explanation of how spatial regulation of pattern is accomplished is not possible without knowledge about regularities by which size of the regulating structure and its inner proportions are defined and the precision and limitations of these regularities. If our understanding of pattern formation is ever to progress beyond a conceptual framework it seems essential that regulative responses of patterns be analysed exactly.

From studies performed on the ciliate Paraurostyla it is evident that experimental alterations in the cell dimensions or deletion of the ciliary pattern cause replacement of the whole ciliature by a new set. The positions of ciliary primordia as well as the formed structures are regulated in such a way that their final dimensions are adjusted to the size of the whole.

The oral apparatus of P. weissei responds to overall dimensions in the following way: all components, that is frontal and ventral membranelles, inner and outer preoral membranelles, show reduction in the numbers of their constituent kinetosomes. This already indicates that during early stages of formation of the ciliary primordia, smaller numbers of kinetosomes are produced.

Reduction in overall number of kinetosomes is not the only response to size diminution - the other is the reduction of the number of serially repeated membranelles (AZM). Frontal as well as ventral membranelles are eliminated in such a way that their respective quantitative ratios remain size invariant. Those constituents of the oral ciliature which exist as single structures (inner and outer preoral membranelles) are never eliminated, they become shorter and the OPM becomes narrower.

All kinetosomes, even from very tiny cells, show normal fine structure - they are composed of 9 triplets of microtubules. The question of whether a kinetosome is the smallest unit of regulation, or whether the number of microtubular protofilaments (Burton, Hinkley & Pierson, 1975) can also vary is not resolved.

Interestingly the cytogeometry within individual adoral membranelles also turned out to be regulated. Within a large range of sizes the internal architecture of membranelles shows size invariant proportionality (Table 3). All these observations taken together suggest that there must be some global control which monitors the overall cell size and the ‘bigger’ parts of the oral ciliature (frontal vs. ventral membranelles and UM’s) and also a local control which monitors the inner proportions of each membranelle in concert with the whole oral apparatus dimensions.

As to the mechanism for achieving such regulation, where everything can change except for the relative proportions, we have no suggestions to present.

We can ask the question: when during morphogenesis does the size adjustment take place? From studies on the formation of the oral ciliature (OP) in P. weissei (Jerka-Dziadosz & Frankel, 1969; Jerka-Dziadosz, 1980ó) and other hypotrich ciliates as well (Grimes, 1972; Bohatier, 1979) we know that the OP originates as a longitudinal field of randomly but densely packed kinetosomes. The length of that field seems to be related to the size of a fragment. This field then segments into promembranelles composed of two equally long kinetosomal rows (R-l and R-2). This segmentation procedes in a wave-front mode from anterior to posterior, similarly to segmentation of the mesodermal streak into somites in vertebrate embryos (Flint et al. 1978; Pearson & Elsdale, 1979). During that process the final number of all membranelles (frontal plus ventral) is established. After that, again in antero-posterior sequence, new rounds of kinetosome proliferation occur anteriorly to each promembranelle and the third (R-3) and then fourth (R-4) rows of kinetosomes are induced at appropriate locations. The process of differentiation of membranelles is accompanied by a shift of the whole developing primordium anteriorly, and then the anterior 1 /4 of the original OP forms the frontal membranelles, oriented at right angles to the long axis of the cell. When the decision concerning differentiation toward frontal-type membranelles is taken is not clear. It seems that the final length of R-l and R-2 of frontal membranelles is set before the membranelles start to move to the right. The proliferation of R-3 and R-4 of each membranelle also starts before shifting of these membranelles. All details of these morphogenetic phenomena will be given in the next publication of this series (JerkaDziadosz, 1980 b).

The results of morphometric and statistical analysis presented in this paper, as well as the analysis of the mode of differentiation of oral structures in P. weissei, seem to confirm the notion about the dual control of pattern formation in ciliated protozoa (reviewed in Frankel, 1979; Jerka-Dziadosz, 1979). This dual control involves so called short-range, or nearest neighbour interaction and a long-range, gradient-field type interaction. The short-range control (basically non-regulatable) is based on restricted self-assembly (Tucker, 1977) where the local environment provides all spatial information required for ordered assembly of kinetosomes and other associated differentiations (Sonneborn, 1975). This mechanism then may partially be responsible for patterning of each individual membranelle.

The long-range forces responsible for asymmetry of structures (Jerka-Dziadosz & Frankel, 1979), positioning of primordia (Jerka-Dziadosz, 1974) are in many respects similar to the forces presumed to act in a morphogenetic field (Frankel, 1974). The simplest and most popular type of model based on Wolpert’s (1969) theory of positional information is that of chemical diffusible components forming a monotonic gradient, whose local concentrations are ‘read off’ and ‘interpreted’ into differentiating structure. These concepts however, possess their own shortcomings because they cannot be adopted either to cases where patterns depend on cell number (Held, 1979), or to those where serially repeating structures such as the vertebrate somites are formed (Cooke, 1975 a, b). A modification of the model which partially accounts for generation and regulation of serially repeated structures was given by Flint et al. (1978) who studied somitogenesis in normal and amputated mutants of mouse. They proposed a wave-form gradient of morphogen which accounts for the definition of somite boundaries and the capacity for regulation in adjusting constant somite number to overall axial length.

Cooke & Zeeman (1976) formulated a theoretical model of pattern formation (though see criticisms of Pearson & McLaren, 1977, and Zahler & Sussman, 1977) which shows how a wave of change passing through a field of coordinate cells can function like an escapement mechanism in a clock to provide for a repetitive series of discontinuities events. This ‘clock and wave-front’model postulated that all the presomite cells become physiologically phase-linked with respect to some oscillator whose rhythm interacts with a wave-front of rapid cell change, causing spatially periodic delays and advances in the overt expression of that wave as cell behaviour. This underlies the spatially regular causes of fissures of cellular de-adhesion segregating successive somites. This model induced new experiments (Cooke, 1978; Elsdale & Pearson, 1979; Pearson & Elsdale, 1979) which showed that a wave is indeed involved in the specification and regulation of segmental pattern.

Since the formation of adoral membranelles from oral primordium in Paraurostyla is essentially a segmentation of uniform longitudinal kinetosomal field into successive subpopulations of kinetosomes (so is the formation of the somatic cirri - Jerka-Dziadosz, 1980a ; Bakowska & Jerka-Dziadosz, 1979) in spatial and temporal sequence down the body axis, it is possible that a determination wave precedes the visible pattern formation. This wave-like change induces the kinetosomes to assemble and aggregate into orderly patterned membranelles.

The difficulty with adopting the above mentioned models to pattern specification and regulation in serially repeated structures in ciliates is that the models require discrete subunits of the field (the cells) to act autonomously but coordinately (cellular processing). If, however, one considers the kinetosomes as units of regulation (in which timing is encoded), and the cell surface structures as a carrier of the positional information variables (Kaczanowska, 1974; Hewitt, 1978), then the general idea of the two-step pattern determination where, first, a presumptive primordial region is programmed and later undergoes segmentation, could be adopted and further tested, This problem will be considered with more detail in the next publication of this series.

The authors would like to thank Dr Joseph Frankel for suggesting some of the statistical tests. We also wish to express our appreciation for valuable comments and discussions concerning earlier drafts of the manuscript by Drs Krystyna Golinska, Janina Kaczanowska and Joseph Frankel. Dr Jonathon Cooke kindly helped us with the English. Excellent technical assistance of Mrs Wiernicka is kindly acknowledged. The work was supported by grant MR.II from Polish Academy of Sciences.

BKowska
,
J.
&
Jerka-Dziadosz
,
M.
(
1978
).
Ultrastructural analysis of the infraciliature of the oral apparatus in Paraurostyla weissei (Hypotricha)
.
Acta Protozool
.
17
,
285
301
.
Bakowska
,
J.
&
Jerka-Dziadosz
,
M.
(
1979
).
Ultrastructural aspect of size dependent regulation of left marginal cirri in a ciliate protozoan Paraurostyla weissei
.
J. Protozool
.
26
, Abstr. no. 252.
Bode
,
H.
,
Berking
,
S.
,
David
,
C. N.
,
Gierer
,
A.
,
Schaller
,
H.
&
Trenkner
,
E.
(
1973
).
Quantitative analysis of cell types during growth and morphogenesis in Hydra
.
Wilhelm Roux Arch. EntwMech. Org
.
171
,
269
285
.
Bode
,
H. R.
,
Flick
,
K. M.
&
Bode
,
P. M.
(
1977
).
Constraints on the relative sizes of the cell populations in Hydra attenuata
.
J. Cell Sci
.
24
,
31
50
.
Bohatier
,
J.
(
1979
).
Morphogenese de regeneration chez le cilie Condylostoma magnum (Spiegel): etude ultrastructurale
.
J. Protozool
.
26
,
404
414
.
Burton
,
P. R.
,
Hinkley
,
R. E.
&
Pierson
,
G. B.
(
1975
).
Tannic acid stained microtubules with 12, 13 and 15 protofilaments
.
J. Cell Biol
.
65
,
227
233
.
Cooke
,
J.
(
1975a
).
The emergence and regulation of spatial organization in early animal development
.
Ann. Rev. Biophys. Bioeng
.
4
,
185
217
.
Cooke
,
J.
(
1975b
).
Control of somite number during development of a vertebrate Xenopus laevis
.
Nature
254
,
196
199
.
Cooke
,
J.
(
1978
).
Somite abnormalities caused by short heat shocks to pre-neurula stages of Xenopus laevis
.
J. Embryol. exp. Morph
.
45
,
283
294
.
Cooke
,
J.
&
Zeeman
,
E. C.
(
1976
).
A clock and wavefront model for control of the number of repeated structures during animal morphogenesis
.
J. theor. Biol
.
58
,
455
476
.
Dembowska
,
W. S.
(
1938
).
Korperreorganisation von Stylonychia mytilus beim Hungern
.
Arch. Protistenkd
.
81
,
89
105
.
Elsdale
,
T.
&
Pearson
,
M.
(
1979
).
Somitogenesis in amphibia. II. Origins in early embryogenesis of two factors involved in somite specification
.
J. Embryol. exp. Morph
.
53
,
245
267
.
Farnsworth
,
P. A.
(
1975
).
Proportionality in the pattern of differentiation of the cellular slime mould Dictyostelium discoideum and the time of its determination
.
J. Embryol. exp. Morph
.
33
,
869
877
.
Finn
,
J.
(
1974
).
A General Method for Multivariate Analysis
.
New York
:
Holt, Rinehart & Winston
.
Flint
,
O. P.
,
Ede
,
D. A.
,
Wilby
,
O. K.
&
Proctor
,
J.
(
1978
).
Control of somite number in normal and amputated mutant mouse embryos: an experimental and theoretical analysis
.
J. Embryol. exp. Morph
.
45
,
189
202
.
Frankel
,
J.
(
1974
).
Positional information in unicellular organisms
.
J. theor. Biol
.
47
,
439
481
.
Frankel
,
J.
(
1979
).
An analysis of cell-surface patterning in Tetrahymena
.
In Determinants of Spatial Organization
(ed.
S.
Subtelny
), pp.
215
246
.
New York
:
Academic Press
.
Grimes
,
G. W.
(
1972
).
Cortical structure in nondividing and cortical morphogenesis in dividing Oxytricha fallax
.
J. Protozool
.
19
,
428
445
.
Held
,
L. I. Jr
. (
1979
).
Pattern as a function of cell number and cell size on the second-leg basitarrus of Drosophila
.
Wilhelm Roux’s Arch. devl. Biol
.
187
,
105
127
.
Hewitt
,
J. A.
(
1978
).
Diffusion gradients, membrane receptors and the acquisition of orientation information by cells
.
J. theor. Biol
.
74
,
297
306
.
Jerka-Dziadosz
,
M.
(
1974
).
Cortical development in Urostyla. II. The role of positional information and preformed structures in formation of cortical pattern
.
Acta Protozool
.
12
,
239
274
.
Jerka-Dziadosz
,
M.
(
1976
).
The proportional regulation of cortical structures in a hypotrich ciliate Paraurostyla weissei
.
J. exp. Zool
.
195
,
1
14
.
Jerka-Dziadosz
,
M.
(
1977
).
Temporal coordination and spital autonomy in regulation of ciliary pattern in double forms of a hypotrich ciliate Paraurostyla weissei
.
J. exp, Zool
.
200
,
23
32
.
Jerka-Dziadosz
,
M.
(
1979
).
Control of pattern formation in ciliated protozoa
.
In Cell Compartmentation and Metabolic channeling
, (ed.
L.
Nover
,
F.
Lynen
and
K.
Mothes
), pp.
495
502
.
VEB Amsterdam
:
Gustav Fischer Verlag, Jena and Elsevier
.
Jerka-Dziadosz
,
M.
(
1980a
).
Ultrastructural study on development of the hypotrich ciliate Paraurostyla weissei. I. Formation and morphogenetic movements of ventral ciliary primordia
.
Protistologica (submitted)
.
Jerka-Dziadosz
,
M.
(
1980b
).
Ultrastructural study on development of the hypotrich ciliate Paraurostyla weissei. II
.
Formation of the oral apparatus, and its bearing on problems of ciliate morphogenesis
. (In preparation.)
Jerka-Dziadosz
,
M.
&
Frankel
,
J.
(
1969
).
An analysis of the formation of ciliary primordia in the hypotrichous ciliate Urostyla weissei
.
J. Protozool
.
16
,
612
637
.
Jerka-Dziadosz
,
M.
&
Frankel
,
J.
(
1979
).
A mutant of Tetrahymena thermophila with a partial mirror image duplication of cell surface pattern. I. Analysis of phentoype
.
J. Embryol. exp. Morph
.
49
,
167
202
.
Jerka-Dziadosz
,
M.
&
Golinska
,
K.
(
1977
).
Regulation of ciliary pattern in ciliates
.
J. Protozool
.
24
,
19
26
.
Jerka-Dziadsoz
,
M.
&
Janus
,
I.
(
1975
).
Discontinuity of cortical pattern during total conjugation of a hypotrich ciliate Paraurostyla weissei
.
Acta Protozool
.
13
,
309
333
.
Kaczanowska
,
J.
(
1974
).
The pattern of morphogenetic control in Chilodonella cucullas
.
J. exp. Zool
.
187
,
47
62
.
Morgan
,
T. H.
(
1901
).
Regeneration of proportionate structures in Stentor
.
Biol. Bull. mar. biol. Lab. Woods Hole
,
2
,
311
328
.
Pearson
,
M.
&
Elsdale
,
T.
(
1979
).
Somitogenesis in amphibian embryos. I. Experimental evidence for an interaction between two temporal factors in the specification of somite pattern
.
J. Embryol. exp. Morph
.
51
,
27
50
.
Pearson
,
M. J.
&
Mclaren
,
D. I.
(
1977
).
A criticism of catastrophe modelling of the differentiative process in amphibian development
.
J. theor. Biol
.
69
,
721
734
.
De Puytorac
,
P.
,
Grain
,
J.
&
Rodrigues De Santa Rosa
,
M.
(
1976
).
A propos de 1’ultrastructure corticale du cilie hypotriche Stylonychia mytilus Ehrbg., 1838 : les caractéristiques du cortex buccal adoral et paroral des Polyhymenophora Jankowski 1967
.
Trans. Am. Microsc. Soc
.
95
,
327
345
.
Socal
,
R. R.
&
Rohlf
,
F. J.
(
1969
).
Biometry. The Principles and Practice of Statistics in Biological Research
.
San Francisco
:
W. H. Freeman
.
Sonneborn
,
T. M.
(
1975
).
Positional information and nearest neighbour interactions in relation to spatial patterns in ciliates
.
Année biol
.
14
,
565
584
.
Stenhouse
,
F. O.
&
Williams
,
K. L.
(
1977
).
Patterning in Dictyostelium discoideunv. the proportions of the three differentiated cell types (spore, stalk, and basal disk) in the fruiting body
.
Devl. Biol
.
59
,
140
152
.
Tartar
,
V.
(
1941
).
Intracellular patterns: facts and principles concerning patterns exhibited in the morphogenesis and regeneration of ciliate protozoa
.
Growth
,
5
, (
Suppl
.),
21
40
.
Tartar
,
V.
(
1967
).
Morphogenesis in protozoa
.
In Research in Protozoology
, vol.
2
(ed.
T. T.
Chen
), pp.
1
116
.
Oxford
:
Pergamon Press
.
Tucker
,
J. B.
(
1977
).
Shape and pattern specification during microtubule bundle assembly
.
Nature
266
,
22
26
.
Wolpert
,
L.
(
1969
).
Positional information and the spatial pattern of cellular differentiation
.
J. theor. Biol
.
25
,
1
47
.
Zahler
,
R. S.
&
Sussmann
,
H. J.
(
1977
).
Claims and accomplishments of applied catastrophe theory
.
Nature
269
,
759
763
.