Preparation method of core-shell structured fibrous scaffolds

Abstract

The present invention relates to a preparation method for core-shell structured fibrous scaffolds, and more specifically to preparing a core part composition and a shell part composition that each have different constitutions by adding calcium phosphate cement and a protein, drug of combination thereof to alginate solution, and then inserting the above core part composition and shell part composition to internal and external nozzle of concentric nozzle respectively to inject into calcium ion aqueous solution and thereby hardening them, thus preparing core-shell structured rapidly setting Alg/-TCP scaffolds capable of controllably releasing a protein or drug.

Claims

1. A method for preparation of a core-shell structured fibrous scaffolds capable of releasing a first protein, drug or combination thereof and a second protein, drug or combination thereof different from the first protein, drug or combination thereof, comprising the following steps: adding calcium phosphate cement, and the second protein, drug or combination thereof to alginate solution to prepare a core part composition; adding calcium phosphate cement, and the first protein, drug or combination thereof to alginate solution to prepare a shell part composition; and inserting the core part composition and shell part composition into an inner and outer nozzle of a concentric nozzle respectively; extruding the resultant core part composition and the resultant shell part composition into a calcium ion aqueous solution so as to form the fibrous scaffolds, in which alginate is cross-linked simultaneously with the hardening of calcium phosphate cement, wherein the alginate solution has a concentration of 1 to 10 weight %; the amount of the calcium phosphate cement in the core part is 10 to 75 weight %, and the amount of the calcium phosphate cement in the shell part is 10 to 75 weight %; the amount of the calcium phosphate cement in the core part is different from the amount of the calcium phosphate cement in the shell part to control release speed of the protein, drug or combination thereof differently; and the calcium phosphate cement consists of an aqueous solution comprising phosphate particles and a substance which catalyzes hardening of the cement, wherein the substance which catalyzes hardening of the cement is hydroxyapatite (HA).

2. The method according to claim 1, in which the calcium phosphate is tricalcium phosphate, monocalcium phosphate, tetracalcium phosphate, dicalcium phosphate, or a combination thereof.

3. The method according to claim 1, in which the protein is growth factor, bovine serum albumin, lysozyme or a combination thereof.

4. The method according to claim 3, in which the growth factor is bone formation factor, angiogenesis factor, or a combination thereof.

5. The method according to claim 1, in which the above drug is antibiotics, an anticancer drug, anti-inflammatory drug or a combination thereof.

6. The method according to claim 1, in which the concentration of the calcium ion is 10 mM to 3 M.

7. The method according to claim 1, in which a hardening time of the calcium phosphate cement is 1 to 10 minutes.

8. A core-shell structured fibrous scaffold which is prepared by the method of claim 1.

Description

DESCRIPTION OF DRAWINGS

(1) FIG. 1 shows the structure of Alg and Alg/-TCP scaffolds obtained by dispensing and hardening in the CaCl.sub.2 solution observed with a naked eye (a), optical microscope image (b) (scale bar: 300 m), low magnification SEM image (c) (scale bar: 100 m) and high magnification SEM image (d) (scale bar: 10 m).

(2) FIG. 2 is a DMA result of alginate-based fibrous scaffolds according to different -TCP content, and the change in storage modulus (E) (a) and loss modulus (E) (b) value according to the frequency (0.1-10 Hz), shown as graph.

(3) FIG. 3 shows in graph the release amount of cyt C released from scaffolds during process of crosslinking in various crosslinking conditions, (a) represents the release amount of cyt C due to change in alginate (Alg) and CaCl.sub.2 concentration while using pure alginate (Alg) aqueous solution, and (b) represents release amount, of cyt C due to change in -TCP content.

(4) FIG. 4 shows a feasible mechanism for the initial release of cyt C from Alg-based scaffolds.

(5) FIG. 5 is a result of studying the release of cyt C from the fibrous scaffolds up to day 28. (a) represents a case when Alg scaffolds are prepared using 3% Alg and 150 mM CaCl.sub.2 solution, (b) represents a case when Alq scaffolds are prepared using 3% Alg and 1 M CaCl.sub.2 solution, (c) represents a case when Alg scaffolds are prepared using 5% Alg and 150 mM CaCl.sub.2 solution, (d) represents a case when Alg scaffolds are prepared using 5% Alg and 1 M CaCl.sub.2 solution.

(6) FIG. 6 is a graph showing release behavior of cyt C when -TCP is added to the Alg scaffolds, (a) represents a case when the scaffolds have different -TCP content, and was prepared through 1 min of crosslinking time, and (b) represents a case when the scaffolds have 50 wt % of -TCP, and were crosslinked under varying time.

(7) FIG. 7 is an optical image of core-shell structured fibrous scaffolds prepared through usage of concentric nozzle. (a) represents a case wherein the core part consisted of pure Alg core part and the shell part consisted of pure Alg, (b) represents a case wherein the core part consisted of Alg and the shell part, consisted of 10 weight % -TCP, (c) represents a case wherein the core part consisted of 50 weight % -TCP and the shell part consisted of pure Alg, and (d) represents a case wherein the core part consisted of 50 weight % -TCP and the shell part consisted of 10 weight % -TCP.

(8) FIG. 8 is the DMA result of core-shell scaffolds with other compositions, showing in graph the change in storage modulus (E) (a) and loss modulus (E) (b) value according to frequency (0.1-10 Hz).

(9) FIG. 9a is a graph showing the result acquired from observing the release behavior after using Alg core-shell scaffolds and loading cyt C into any one of core part or shell part.

(10) FIG. 9b is a graph showing the result acquired from observing the cyt C release after changing the composition of any one of the core part or shell part while changing the amount of -TCP.

(11) FIG. 9c is a graph showing the result acquired from observing the cyt C release after changing the composition of any one of the shell part or core part when cyt C is loaded into the core part.

BEST MODE

(12) The present invention is explained in more detailed through the Examples below. However, these Examples are merely intended to illustrate the present, invention and are not intended to limit the scope of protection.

Example 1 Materials and Methods

(13) Materials and Fibrous Scaffold Deposition

(14) -TCP was obtained by sintering a mixture of calcium hydrogen phosphate (CaHPO.sub.4, Sigma-Aldrich, C7263) and calcium carbonate (CaCO.sub.3, Sigma-Aldrich, 239216) at 1400 C. and subsequent quenching. The sintered powder was milled in a planetary mill and added with 2 wt % hydroxyapatite (HA) crystallites as a seed for the phase-trans format, ion of cement from -TCP into HA. The -TCP had a median particle size of 5.2 m as determined from laser diffraction (Malvern, APA5001SR).

(15) Na-Alg (Sigma-Aldrich, A2158) was dissolved in water at two different, concentrations (3 or 5%). The -TCP powder was mixed with the Alg solution at ratios of 10, 50, or 75 wt % of -TCP with respect to the composite. For the scaffolding process, a dispensing machine (KD Scientific) was used. The mixture was dispensed into a fiber form through a 23G syringe at an injection rate of 50 mL/h into a bath containing CaCl.sub.2 (150 mM or 1 M in distilled water), during which the fibrous scaffold form was preserved without being disintegrated by the crosslinking reaction of Alg with calcium ions. After dispensing 0.5 mL of the solution, the dispensed scaffolds were left in the CaCl.sub.2 solution for different times to better conduct crosslinking (crosslinking time: 1, 5, 10, and 30 min). Afterward, the fibers were shaped in a Teflon mould with dimensions of 12 mm diameter6 mm height.

(16) Design of Core-Shell Scaffolds

(17) A concentric nozzle (inner 17G and outer 23G) was specifically designed and used to produce core-shell-structured fibrous scaffolds. Each solution with specific composition (Alg or Alg/-TCP) as described in the below Table 1 was separately fed into the outer and inner syringes. The compositions of the core-shell were varied as summarized in Table I. Each syringe was attached to an injection pump connected through a microtube. Core-shell-structured Alg fiber was then injected at a rate of 50 mL/h through the coconcentric nozzle within a bath containing 150 mM CaCl.sub.2 solution for 1 min.

(18) TABLE-US-00001 TABLE 1 Case in which protein.sup.1) is Case in which protein.sup.1) is comprises in the shell part.sup.2) comprised in the core part.sup.2) -TCP content of -TCP content of -TCP content of -TCP content Core part Shell part Core part of Shell part (wt %) (wt %) (wt %) (wt %) 0 0 0 0 0 10 0 10 50 0 [Note] .sup.1)cytochrome C(cyt C)was used as model protein .sup.2)protein is designed to be comprised in either the shell part or the core part

(19) Characterization and Mechanical Tests

(20) Scanning electron microscopy (SEM) was carried out using a JEOL JSM-6510 apparatus to investigate the microstructure of the scaffolds. The samples were sputtered with platinum for the SEM examination. The mechanical properties of the scaffolds (12 mm diameter6 mm height) were measured by dynamic mechanical analysis (DMA; MetraVib, DMA25N) in the parallel plate configuration. Mechanical spectrometry was carried out using dynamic frequency sweep with frequencies ranging from 0.1 to 10 Hz at 37 C. and with strain amplitude of 5%, which was in the linear region of viscoelasticity. Both autotension and autostrain adjustments were applied. Force was ramped from 0.001 to 0.2 N, and the maximum allowed strain was set at 10%. The storage modulus and loss modulus (E) of the samples were measured.

(21) Protein Loading and Release from the Scaffolds

(22) First, the capacity to in situ load proteins within the fibrous scaffolds during the dispensing process was observed.

(23) Cytochrome C (cyt C) was used as the model protein, reflecting its common use as the model for growth factors. 500 g of cyt C was added to 0.5 mL of solution (either Alg or Alg/-TCP composites), which was then dispensed into fibrous scaffolds in CaCl.sub.2 solution. For the case of core-shell scaffolds, cyt C loading was designed to be within only the inner or outer part, and the effect of the compositional change (inner or outer composition) on the release of protein was investigated. After crosslinking the scaffolds for different time points (1, 5, 10, and 30 min), the amount of cyt C was analyzed from the supernatant to detect the loading quantity. The amount of cyt C released was measured using a Libra S22 spectrophotometer at an absorbance 408 nm (Biochrom). Loading experiments continued by changing other crosslinking parameters, including concentration of Alg (3 and 5%) and CaCl.sub.2 (150 mM and 1 M). Furthermore, the effect of -TCP amount in the mixture solution with Alg (0-75 wt % of -TCP) was also investigated. To study the release of cyt. C from the scaffolds, each scaffold sample was immersed in 1 mL phosphate buffered saline (PBS) for different periods of up to 42 days. The cyt C released was assessed spectrophotometrically at an absorbance of 408 nm using the aforementioned Libra S22 apparatus. The release quantity was interpreted after normalized to the loaded quantity. The medium was refreshed at each time point of the assay.

Example 2 Characteristics of Deposited Fibrous Scaffolds

(24) The macroscopic morphology of the Alg and Alg/-TCP scaffolds obtained by means of dispensing and hardening in CaCl.sub.2 solution is shown in FIG. 1. The addition of -TCP powders rendered the scaffold opaque. An extensive fibrous network with highly macroporous structure developed for all scaffolds, which was expected to allow cell migration and penetration (Lee G S et al., Acta Biomater, 2011, 7, 3178-3186). Optical images of the composite scaffolds revealed the dispersion of -TCP particles (several micrometers in size) in the Alg matrix. The sizes of the fibers and the interspacings (macropores) of the scaffolds with different compositions were similarly observed; fibers of 200-250 m and macropores of 200-500 m. SEM images also revealed -TCP particles distributed in the Alg phase. When the amount of -TCP was high (75%), the Alg fibers were completely covered with -TCP. A similar morphological feature has also been observed in the collagen/-TCP scaffold (Perez R et al., J Mater Sci Mater Med, 2011, 22, 887-897). It should be borne in mind that -TCP transforms to calcium-deficient HA under physiological conditions, leading to phases consisting of Alg and HA (Lee G S et al., Acta Biomater, 2011, 7, 3178-3186; Perez R et al., J Mater Sci Mater Med, 2011, 22, 387-897). As the -TCP-containing composites retain the biomimetic carbonated apatite phase, they are considered to provide beneficial matrix conditions in terms of chemical composition for cells to engage in osteogenesis and to function in bone formation.

(25) The effect of the -TCP incorporation (10, 50, and 75%) into scaffolds on the mechanical properties was investigated under dynamic conditions using DMA. Storage modulus (E) and loss modulus (E) values were recorded as a function of frequency (0.1-10 Hz) as shown in FIG. 2. For all cases, the E values were higher than the E values, demonstrating that the fibrous scaffolds behaved in a more elastic fashion than as a viscous network. The higher the amount of -TCP added, the higher were the mechanical properties (both E and E), with the E change being especially pronounced, suggesting the stiffening role of -TCP phase in the polymeric Alg. The E values of the Alg/-TCP composite scaffolds were about 150, 200, and 800 kPa with 10, 50, and 75% -TCP addition, respectively, which were significantly enhanced values with respect to that of Alg scaffold (80 kPa). The stiffness of the scaffolds could be tuned with the addition of -TCP. It is worth noting that for the case of 75% -TCP/Alg composite scaffold, the improvement was as high as 10 times.

Example 3 Loading of Cyt C and Crosslinking of Scaffolds

(26) Having confirmed the beneficial properties of the Alg/-TCP composite scaffolds in the above Example 2, it was investigated the capacity of the scaffolds in loading and delivering therapeutic molecules, particularly growth factors.

(27) Cyt C was used as the model protein. It has been frequently used to represent the behavior of growth factors like fibroblast growth factors owing to the similarity in size and charge characteristic. First, it was observed the loading behavior of cyt C within the scaffolds. Cyt C was loaded in situ within the material solution, which was then allowed to harden in highly concentrated CaCl.sub.2 solution due to the effective crosslinking of Alg phase and scaffold. Thus, it was possible that the crosslinking step influences the loading behavior of cyt C. In fact, a preliminary study established that a large portion of cyt C initially used was released during the crosslinking process. Therefore, in this example, it was analyzed systematically the loading behavior of cyt C in the crosslinking step by varying the crosslinking conditions, such as CaCl.sub.2 concentration (150 mM or 1 M), Alg concentration (3 or 5%), and crosslinking time (1, 5, 10, and 30 min). The effect of divalent ions on the Alg crosslinking has been previously studied. Ca.sup.2+ produced better performance than other cations including Ba.sup.2+ and Sr.sup.2+ (Acarturk S T, J. Microencapsul., 1999, 16, 275-290). The Ca.sup.2+ ions replace Na.sup.+ ions in Alg structure and make strong bonds with Alg through ionic interactions, forming a crosslinked network. Therefore, it was considered that the ion concentration and Alg concentration (3 and 5%) should importantly determine the crosslinking process and the resultant Alg-Ca networks. Increasing the Ca.sup.2+ ions-would be helpful for more crosslinks, whereas certain optimal Alg concentrations may exist to achieve rapid crosslinking as well as more highly crosslinked structure. Moreover, a certain time period may be required to complete crosslinks.

(28) The release quantity of cyt C released from the scaffolds during the crosslinking process at varying crosslinking conditions is shown in FIG. 3(a). Alg scaffold was used for this study. It was clear that the crosslinking time affected the cyt. C release most significantly although there were some differences in the cyt C release depending on the other parameters (Alg or CaCl.sub.2 concentration). Cyt C release continued to increase with increasing time. Specifically, 5-7% at 1 min became 22-43% at 30 min.

(29) The effect of -TCP addition on the cyt C release was also investigated (FIG. 3(b)). A similar cyt C release behavior was observed in all the -TCP-added composite scaffolds, with no clear effect of the amount of -TCP: an on-going cyt C release with crosslinking time. Therefore, it is concluded that the release of cyt C protein is primarily dependent on the crosslinking time. Thus, shortening the crosslinking time is inevitable to reduce the protein release.

(30) The possible mechanism of the initial release of cyt C from the Alg-based scaffolds is illustrated in FIG. 4. Compared to crosslinking of pure Alg [FIG. 4(a)], when cyt C protein was pooled in the Alg solution [FIG. 4 (b)], it can bind to Alg molecules through weak interactions as cyt C has a net positive charge, whereas Alg is negatively charged. However, when immersed in CaCl.sub.2 solution, Ca.sup.2+ ions bind Alg molecules more strongly, repelling cyt C molecules, which results in initial cyt C release. When -TCP powders are added to Alg solution [FIG. 4(c)], there should also be interactions of cyt C with -TCP, which however, also are not enough to overcome the ion exchange by Ca.sup.2+ and cyt C release out from the scaffolds during the crosslinking process. Storing the scaffolds in a highly concentrated CaCl.sub.2 cross linking solution will cause high influx of Ca ions, which drives the release of cyt C molecules from the scaffolds. It is also considered that the cyt C molecules pooled in the Alg/-TCP solution could not have sufficient time to be tightly adsorbed on the surface of -TCP particles. Therefore, the weakly bound cyt C molecules, particularly those present at the outer surface of the fibers, should be released rather rapidly at the beginning stage of crosslinking. Thus, combining cyt C with -TCP powders could initially possibly help lessen the rapid cyt C release. Moreover, it should also be noted that the -TCP state transforms to an apatite state with time, which will affect the release of cyt C molecules at a much later stage [FIG. 4 (c)].

Example 4 Prolonged Release of Cyt C from the Scaffolds

(31) The release of cyt C from the fibrous scaffolds was examined in PBS at 37 C. for periods up to 28 days. Tests were made at different cross-linking conditions and results were plotted considering the cyt C release amount, as the starting point of each graph (FIG. 5). For the Alg scaffolds prepared with 3% Alg and under 150 mM CaCl.sub.2 solution [FIG. 5(a)], crosslinking times over 5 min resulted in high initial burst. (15% for 1 day) and almost, saturation in the release, however, a short crosslinking time of 1 min showed a slow and steady release of cyt C over the period of 28 days although the total release amount was 18%. In the case of higher Alg concentration (5%), the scaffolds prepared at both CaCl.sub.2 concentrations showed the initial burst effect and a saturation behavior without regard to the crosslinking time (FIG. 5c and FIG. 5d). Based on the results, the crosslinking of 3% Alg in 150 mM CaCl.sub.2 for 1 min showed only the possible sustainable and continual release profile of cyt C, which however, exhibited a very limited amount of cyt C released (18%), even after 28 days.

(32) The effects of -TCP addition to the Alg scaffolds on the cyt C release behaviors were also observed as shown in FIG. 6. The composite scaffolds crosslinked for the short time of 1 minute [FIG. 6(a)] profiled a continual release of cyt C, and the addition of -TCP significantly enhanced the release rate with time. Consequently, after 42 days (6 weeks), the total release amounts were 35, 48, and 80% for 10, 50, and 75 wt % -TCP, respectively (higher than 18% for pure -TCP), given that the cyt C release amount was small and almost similar (5-7%). The 50% -TCP/Alg scaffolds obtained with different crosslinking time exhibited similar on-going release profile with time, without showing a plateau, being different from those observed in pure Alg [FIG. 6(b)], In consideration of the initial cyt C release amounts of each case (higher release corresponding with longer crosslinking time), the consequent cyt C release amounts were 48% (1 min), 25% (5 min), 35% (10 min), and 25% (30 min), demonstrating different release amounts with respect to crosslinking time. Taking all the release profiles of cyt C from the -TCP-added Alg scaffolds into account, the noteworthy finding was the continual (much like zero-order) and sustainable (still releasing after 6 weeks) release profiles without regard to the crosslinking time (and, thus, initial cyt C release), which contrasted to the cases observed in pure Alg scaffolds. These facts support the potential usefulness of the -TCP/Alg composite scaffolds for continual long-term delivery of growth factors. Furthermore, the big difference in the release rate of the scaffolds achieved by varying the -TCP amount should be properly utilized in applying the system for delivery of growth factors at a controllable rate and quantity. Thus, the core-shell-structured scaffolds were designed to have the position of the composition of the core or shell part adjusted, to thereby incorporate growth factors separately and tailor their sequential release.

Example 5 Core-Shell Structured Scaffolds and Cyt C Release

(33) FIG. 7 shows the optical image of the core-shell-structured fibrous scaffolds produced by using a concentric nozzle. Different compositions were used to comprise the core and shell parts. Owing to the differing amounts of -TCP powders added to the scaffold, the core-shell structure was optically discernible. As the same injection speed was used for both core and shell, the volume of core and shell was equal, and the measurement of volume from the two-dimensional images (core diameter of 1400 m and shell thickness of 300 m, and taking cylindrical shape) also gave a similar result.

(34) The DMA of the core-shell scaffolds with different compositions was also performed as shown in FIG. 8. The Alg of the core-shell-structure showed storage modulus (E10-20 kPa and loss modulus (E) 15-20 kPa, which were smaller than the values in single Alg fiber (vs. FIG. 2), and this may be due to the difference in whole fiber size (650 m for single fiber vs. 2000 m for dual fiber) and the consequent macroporosity between fibers. The addition of -TCP either to core or to shell increased both modulus values and the more so in E values, suggesting a more significant role in elastic stiffening of the Alg-based scaffolds. This improvement in modulus values with -TCP addition was similarly observed in single fibers. As a whole, the modulus values were higher in the scaffold when the total amount of -TCP present in the core and shell part was higher: 0%<10% shell<50% core<50% core+10% shell<75% core<75% core+10% shell<75% core+50% shell. Again, the mechanical stiffness values, largely mimicking native bone, of the -TCP-added scaffolds with core-shell structure in comparison to the Alg scaffolds, are of great merit for their use in bone tissue engineering applications.

(35) Further, the cyt C release behavior from the core-shell-structured scaffolds was investigated. Cyt C was loaded in either the shell or the core part while varying the composition of each part. It was thought that the cyt C loaded in the core would show a delayed release profile owing to the outer shell layer, being compared to the case loaded in the shell, which would release more quickly. Moreover, varying the composition (-TCP/Alg amount) of either core or shell will alter the cyt C release profile.

(36) First, with Alg core-shell scaffold [FIG. 9(a)], cyt C was loaded in either core or shell part, and the release behavior was observed. As expected, the cyt C release from the core showed a delayed pattern with respect, to that released from the shell. Results demonstrated clearly that by switching the loading of cyt C in either core or shell it was possible to control the release profile of cyt C.

(37) Next, the composition of either the core or the shell was changed by varying the -TCP amount and then the cyt C release was observed. When cyt C was loaded into the shell [FIG. 9(b)], the compositional variation (Alg vs. 10% -TCP) showed different release profiles: lower initial burst and an on-going release profile with TCP addition, as was similarly observed in the case of single fiber scaffolds (FIG. 6).

(38) When cyt. C was loaded in the core [FIG. 9(c)], changing the composition of either shell or core part also affected the protein release. At the core composition of pure Alg, cyt C release from the core only slightly changed depending on the shell composition (FIG. 9c green line (-.box-tangle-solidup.-) vs blue line (-.diamond-solid.-). When the core composition was 50% -TCP/Aig, the cyt. C release from the core was substantially enhanced with respect to the cases with Alg in the core (FIG. 9c, red line (-.square-solid.-) and purple line (--)). Furthermore, the release was significantly higher when the core part, composition was 50% -TCP/Alg then when shell part contained 10% -TCP (FIG. 9c purple line (--) vs red line (-.square-solid.-). In particular, for this case, the total cyt C release amounts at 42 days were as high as 80-100%, and continual releases were achieved 5 days onward.

(39) The results of cyt C release from the core-shell-structured scaffolds support the design of a scaffold that can carry dual growth factors, where the rapid releasing factor is placed at the shell part while the one requiring more sustainable profile at the core part. This is easily implementable just by positioning each factor at different parts during the deposition process, and is primarily the benefit of the crosslinking nature of the Alg scaffold in divalent ionic solution. Although the process is conducted under water-based (nontoxic solvent) and mild temperature conditions, the high ionic concentration required to crosslink Alg resulted in some release (leaching out) of proteins. A possible way to preserve a major amount of proteins was to shorten the crosslinking time (i.e., by several minutes), and this was a universal phenomenon without regard to the scaffold compositions. As to the effects of -TCP, it can be envisaged that tuning the release rate can be more ambitious when the -TCP was introduced into Alg scaffold. Adding the -TCP up to 75% enabled the scaffolds to release cyt C continuously at relatively high quantity for a long period (over 6 weeks), which potentiates the capacity of the dual-structured -TCP/Alg composite scaffolds in loading and longterm delivery of growth factors.

(40) Using specific growth factors targeting bone regeneration, such as osteogenic factors (e.g., BMPs) and angiogenic factors (e.g., VEGF) to position each part of the core-shell structure with proper compositions, is one example. The loading of VEGF in the shell while positioning BMP in the core part of Alg will get the profile of sequentially released VEGF and BMP, for all the compositions chosen as deduced from FIG. 9(a). Furthermore, altering the composition of each part also controls the sequential release rate of the two different growth factors. Specifically, using 0% -TCP sustains more the release of BMP placed in the core than using 10% -TCP as deduced from FIG. 9(c).