Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
9 S 3
Some years ago, the time-consuming and rather burdensome
procedure of serially dlluting bacteria was improved upon considerably by
depositlng the bacteria-laden llquid ln a spiral pattern on the
previously-plated agar surface of a single rotating culture difih. No
S useful purpose would be served by going into detail concerning this
well-known technique or the method and apparatus used for making such
depositions or counting the colonies of bacter~a thus deposited since these
matters are fully set forth in U.S. Patents Nos. 3,799,844; 3,892,632; and
3,962,040, all of which have been assigned to the Department of Health
~ducation and Welfare of the United States. For the present it should
suffice to point out that the bacteria-laden liquid was laid down in the
form of an Archimedes spiral upon the agar coating so as to effectively
produce a continuous dilution thereof starting near the center with a high
concentration and ending near the periphery of the dish with a low one.
In the te6ting for chemical mutagens to obtain a so-called
"dose-response curve", the conventional practice used to be that of mixing
sensltlve bacterla with several different concentrations of chemical in a
liquid and then platlnp each concentration on the agar surface of a
separate culture dish. The several culture dlshes were then incubated and
the number of colonies on each plate counted. A direct correlation existed
between the number of colonies present and the concentration of the
chemical present which relationship forms the basis for the dose-response
curve. More recently, this outdated procedure has been altered by spirally
plating the test chemical using the above-mentioned apparatus and
' 25 . overlaying with sensitive bacteria. A complete dose-response relatlonship
ls obtained ln one culture dish by so doing.
Now, ln order for this spiral-depositlon pattern to be
quantitatively useful, some means must be provided for reading the number
of colonies present ln a given area. For lnstance, the number of colonies
- 2 -
., .
~7~953
may be low at the beginning of the deposltlon because the high concentration
of the chemical at this point may prove toxlc to the bacterla. On the other
hand, toward the end where the chemical concentration is dllute, this same
dimunltion in colony count can occur, and oftentimes does, because the amount
of chemical present is too small to be effective. In this situation, at
some point in between the number of colonies reaches a maximum per unit
area and this translates into the optimum chemical concentration.
So far as applicants are aware, only one so-called "counting
grld" is available which, upon being placed beneath the completed spiral
is effective to quantify the number of colonies present withln a prescribed
area thereof. This grid is marketed by Spiral Systems, Inc., oi Bethesda,
Maryland and it consists of a circular, not a spiral, pattern divided into
five major concentric rings. These rings are each then subdivided radially
into a total of eight wedges or sectors.
In performing a standard assay using such a grid, the to~al
volume of bacteria-laden liquid deposited in each of the discrete sectors is
known, the several such sectors carrying distinctive number or letter
designatlons. The number of colonies present in each of the several sectors
is counted until a statistically slgnificant number is found; whereupon, the
concentration of cells m the original sample can be calculated based upon
the known volume of bacteria-laden liquid present in the particular sector.
Using the foregoing technique for a standard assay, the starting
and stopplng points for the deposition are not important. Instead, the
plate is plsced randomly with respect to these counting sectors.
The provision of a dose-response curve, on the other hand, lnvolves
a more complex procedure for which the prior art concentric-circle grid
proved to be unsatisfactory. When an attempt was made to count the colonies
,, .
1 17~953
sequentlally starting wlth the beglnnlng ~f the splral d~posltion pat~ern,
the resultlng co~mts exhiblted periodlclty rather than continuous change.
More speclflcally, the counts withln each rlng showed regularly-occurring
highs and lows as the number of cells wlthln each wedge-shaped sector was
counted uslng what is known in mlcrobiological clrcles as the "Ames assay."
Part of the problem appeared to be caused by the fact that the
prlor art grld was circular and did not, therefore, follow the spiral
deposition pattern along which the sample was laid down. Unfortunately,
merely using a grid or template which did, in fact, follow the spiral
deposition pattern also proved to be unsatisfactory for several reasons,
the most significant of which is the fact that it proved to be impractical
to count continuously along the exact deposition lines of the spiral. One
reason for thls is that the chemical diffuses into the convolutions of the
splral on elther side thereof thus introducing inaccuracies in any count
predlcated upon the assumptlon that each such convolutlon remalns separate
and dlstlnct.
Whlle this dlffusion problem was formldable ln itself, even more
so was that of overcoming the periodicity exhlbited when the concentric
clrcle grld was used, These problems along with that of determining the
sl~e, length and arrangement of counting segments which would produce a
reliable cell count presented applicants with a severe challenge, one which
they were fortunately able to solve by the slmple yet unobvlous expedient
of flrst lncluding within each segment of the countlng grid, not one, but
several deposition tracks. ~ext, while using a spiral grid, it proved to
be necessary that it not follow the depositlon spiral but instead intercept
same at intervals selected to eliminate periodicity. ~inally, not one but
two dlfferent segment patterns have proven effective, the first conslsting
of a plurality of segments of different lengths depending upon their radlal
distance from the center and all lying within a common clrcular sector
located opposite the point where the spiral deposition commenced and a
second where all but the final segment have the same length measured along
the grid spiral.
It is, therefore, the principal object of the present invention
to provide a novel and improved spiral-patterned counting template for use
in counting cell colonies produced in response to a chemical laid down in
accordance with an Archimedes spiral.
A second object is the provision of a template of the type
aforementioned which elimininates the problems associated with diffusion
from one convolution of the deposition spiral to the next.
Another object of the invention herein disclosed and claimed
is the provision of a segmented cell-counting grid which integrates with
an Archimedes deposition spiral so as to eliminate periodicity.
Still another object of the within described invention is that
of providing a novel spiral grid which is adaptable for use with automated
counting procedures.
An additional objective is the provision of a device of the
type aforementioned which is ideally suited for use in obtaining the data
required for complete microbial dose-response curves in a single culture dish.
Further objects are to provide a spiral counting template
which is simple, easy to use, versatile, inexpensive, reliable and fully
compatible with existing spiral plating machines.
Broadly stated, the invention provides, for use in combination
with a spirally-deposited solution containing steadily decreasing con-
centrations of a chemical whose effect upon sensitive bacteria is to be
analyzed, an improved means for counting the latter. The improved counting
means comprises a transparent underlay imprinted with indicia defining
a spiral grid pattern having a beginning point and an end point closely
4953
approximating those of the deposition spiral when superimposed thereupon,
said grid having at least two convolutions defined in terms of the inner
and outer marginal edges thereof and having a clear space therebetween,
adjacent covolutions sharing a common marginal edge, the width of the
convolutions being constant and spanning at least three convolutions of
the deposition spiral, the rate at which said grid convolutions diverge
radially from the starting point thereof being greater than the corresponding
rate of the deposition spiral such that the convolutions of the former cut
across the convolutions of the latter at least twice between the beginning
and the end thereof, and at least a portion of the convolutions of said
grid spiral being segmented to define a plurality of discrete areas within
which the bacteria present can be counted.
Other objects will be in part apparent and in part pointed
out specifically hereinafter in connection with the description of the
drawings that follows, and in which:
Figure 1 is a plan view of a template having one form of the
spiral counting grid of the present invention inscribed thereon;
F;gure 2 is a plan view similar to Figure 1 and to the same
scale showing an alternate form of the grid using the identical spiral but
different segmenting thereof, and,
Figure 3 is a diagram to a considerably larger scale showing
only the central portion of the counting grid in relation to a corresponding
portion of the spiral deposition pattern.
-- 6 --
1 174g~3
Referring next to the drawings for a detailed descrlption of the
present invention, both Figs. 1 and 2 show a circular template T of
transparent material having a spiral grid G inscribed thereon. The basic
spiral grid G remains the same in both the Fig. 1 version whlch has been
S broadly designated by reference numeral 10 and in the alternate version
thereof shown in Fig. 2 and which has been similarly identified by reference
numeral lOM. Spiral grid G, in the particular form shown, consists of a
total of five complete 360 convolutions beginning at starting point S
adjacent the center C of the template and terminating at point E on the
periphery thereof, point E being on the same radial line as starting point
S. It can also be seen that the width of the five convolutions of the grid
spiral remain constant from the beginning (S) to the end (~).
Next, while no attempt has been made to illustrate the entire
depositlon pattern D due to space limitations, enough thereof has been shown
by wsy of phantom lines ln Flg. 3 to which reference will now be made to
delineate certain slgnlficant relationships between it and the counting
grid G, the correspondlng portlon of whlch has been shown in full lines.
It will be noted that deposltion splral D consists of a considerably
tighter spiral than its countlng grid counterpart G. Moreover, all of the
convolutlons of the deposltlon spiral, while also of a constant width, are
considerably narrower than the convolutlons of the grld spiral. More
specifically, lt has been found that the width of the counting grid spiral
is advantageously selected such that it totals a whole number multiple
greater than one of the width of the deposition spiral convolutions plus
approximately one-half of such width. As illustrated, the selected whole
number multiple is three which means that with the deposition splral
convolutions each having a width of 0.8 milllmeters, the counting grid
9 5 3
convolutions wlll each be 3.5 X 0.8 or 2.8 cm wide. Looking at this
relatlonship somewhat differently, each countlng grld convolution will
preferably extend across a total of three deposltioD grld spiral
convolutlons and half way across the fourth. It has been found that by so
dolng, the diffuslon effects lnherent 1D the spiral plating technique can
be cancelled out to the point where they are no longer statistically
significant.
The periodicity problem associated with the conventional counting
method was solved by simply dèsigning the counting grid so that it did not
repeat its spatial relatlon to the deposition grid during each 360
excursion, but at st only on alternate convolutions. More speclflcally,
it was dlscovered quite unexpectedly that periodlcity in the count
virtually disappeared when the counting grid diverged at a rate such that
it intersected the correspondlng portion of the deposition pattern at
radially misaligned points except, perhaps, the beginning and the end. The
preferred pattern is the one illustrated,wherein the grid spiral intersects
the deposition spiral precisely five times from the beginning to the end
thereof, the latter polnts being coincident with one another. This works
out such that the counting grid intersects the depositlon grid 2-1/2 tlmes
per re~olution and if one divldes 2-112 into 360, it will be found that the
first intersection is 144 from the start, the second 288, the third 432
or 72 past S, the fourth 216 past S, and the fifth of course, 360 past S
agaln whlch brlngs lt back into radial alignment with the original 0, all
of which is shown in Fig. 3.
Thus, in accordance with the preferred embodiment shown in
Fig. 3, starting at point "0" where both the innermost deposition pattern
edge and the innermost countlng grid line orlglnate, it wlll be seen that
section Gl of the latter intersects the outer edge D2 of the former at a
point X spaced angularly from starting point S preclsely 144. This same
-- 8 --
~ 1749~3
innermosL countlng grld llne (:1 crosses the outside edge of the second
deposltion splral convolution D3 at a point Y 2~8~ from the start S. Thls,
of course, is 72D from the start which means that innermost ~rid line Gl
will only be half way across the space between the inside and outside edges
(D3 and D4) of the fourth deposition splral convolution when lt completes
one full 360~ revolution. Looking at this relationship another way, a
given llne of the counting grid G intersects an edge of the deposition
pattern spiral D exactly 2-1/2 times each revolution. Note also in Fig. 3
that the end of countlng grld line G2 and the beginning of counting grid
line G3 intersect the outside on edge D8 of the seventh convolution of the
deposition spiral at a point in radial alignment with the origin point 0.
The resultant repetitive relationship between angularly-disposed
convolutions of the counting and deposition pattern results in an
intersectlng pattern has, as prevlously noted, proven effective in
eliminating the periodicity which was experienced with the Ames test when
- using the prlor art concentrlc circle grid. Having thus devised a grid
capable of eliminating, or at least controlling to a statistically
lnsignificant degree, the problems of the prior art grid having to do with
diffusion and periodicity. There remained, however, the very significant
problem of how to make the cell colony count with sufficient accuracy to
provide the data needed for a smooth dose-response curve, two different
solutions to this problem having been found, the first of which has been
shown in Fig. 1 to which detailed reference will next be made.
Arranged in diametrical relation to the start S of the counting
grid G is a wedge-shaped sector W bounded by radii cooperating to subtend
an angle of 90. In the particular form shown, the sector extends 45 on
either side of the diameter containing the start S and end E of the counting
grid spiral. The grid spiral, of course, includes five convolutions whlch
is the case along any radius. It should be noted, however, that by placlng
_ g _
1749.53
the scctor W opposite the start and end of the grid as opposed to the same
side thereof, all scgmcnts of the spiral convolutions go all the way from
one sid~ of the sector to the other. Obviously, this same condition can
be found anywhere on the grid except where the radius containing the start
and end thereof is located somewhere within the sector W which would result
in one foreshortened segment either at the beginning or at the end.
In the particular form shown~ sector W is divided into a total of
eight counting segments with the two (1 and 2) nearest the center being a
full 90~ in angular extent while all the remaining ones, 3-8, inclusive,
are only 45 in angular extent but two such segments are provided in each
convolutio~ of the spiral arranged in end-to-end relation so as to subtend a
total of 90. Segment 2 subtends a greater area than l, 3 plus 4 a greater
area than 2, and so on out to the fifth convolution where segments 7 and 8
have a combined area greateT than any of the others. All the divided
segments are divlded half way so that segments 3 and 4 are equal in angular
extent as are 5 and 6 along with 7 and 8. Segment 3 is not, however,
equal iD area to segment 4, the latter being slightly larger due to its
being farther from the center C of the grid. In like manner, starting
with segment 3, all higher numbered segments 4 through 8 are largeT in area
than any preceding one except 2 and 1. The segmented grid just described
results in segments which have a decreasing total volume of liquid per
segment outward from the center as shown in the following table:
TABLE I
Volume per segment for Fig. 1 template
Segment 11 Total ~l/segment
l 1.6
2 1.1
3 0.26
4 0.24
0.23
6 0.22
7 0.20
8 0.10
-- 10 --
~ 1~4953
Each segmcnt, therefore, has a characterlstic total volume of liquid and
since the basis for a dose-response curve is the fact that the cell colonies
must be counted in discrete units of length along the grid spiral so that
the number thereof can be related to this known volume and, more
particularly, to the concentration of chemical contained within this known
volume~ segmented grid of Fig. l provides the required data in a manner
heretofore unknown in the art. The segmented grid shown has no segment under
2 cm in length and this has been found to be quite adequate for purposes of
providlng the cell count data needed for a good dose-response curve.
Next, looking at Fig. 2, a dified template lOM has been shown
which is overlayed by the exact same spiral grid G but which is segmented
somewhat differently than the template lO of Fig. 1. In the former, only
a 90 sector of the grid was segmented, whereas, in the latter, all the
convolutions of the spiral are segmented beginning at the start S and
stopping at the end E. In t.he Fig. 2 embodiment, however, each of the
first fifteen segments is of equal length except for the last or outermost
one (16) which is only half as long as the others. There is the further
proviso that the segments in adjacent convolutions of the spiral be staggered
such that no two, including the flrst and last, terminate along the same
radial line. The first of the sixteen total segments starts at the start
of the spiral deposition pattern and continues with all segments from there
on being arranged in end-to-end relation. In the actual grid used on a
standard lOcm culture dish, segments 1-15, inclusive, are each 5cm in
~ length with se~ment 16 being only 2.5cm long. Thls segmented counting
pattern results in a decreasing volume of material being deposited from
beginning to end that closely approximates that which is shown ln the
following table:
~. 1749~S3
TABLE Il
Volume per Segment for Flg. 2 template
Segment ~ Total ~l/se~ment
1 2.7
2 3.0
3 1.9
4 1.5
1.3
6 0.98
7 0.95
8 0.74
9 0.66
0.49
11 0.42
12 0.42
13 0.38
14 0.32
~ 15 0.27
2~ 16 0.10
For manual counting, the templates T are inscribed on a
transparent materlal (plastlc or glass) which can be illuminated from behind.
The culture dishes are placed over the templates and the starting po$nts
indicated by the arrows in Figs. 1 and 2 are aligned. The colonies in
each segment are then counted in accordance with techn$ques well known in the
art. The counts are recorded, and then related to the total quantity of
chemical deposited in a segment using the known volume in a segment. The
cell number is then plotted against concentration of the chemical to obtain
a dose-response curve.
- 12 -
