Uniform intratumoral distribution of radioactivity produced using two different radioagents, 64Cu-cyclam-RAFT-c(-RGDfK-)4 and 64Cu-ATSM, improves therapeutic efficacy in a small animal tumor model
The present study proposed a new concept for targeted radionuclide therapy (TRT) to improve the intratumoral distribution of radioactivity using two different radiopharmaceuticals. We examined the efficacy of a combination of a tetrameric cyclic Arg-Gly-Asp (cRGD) peptide-based radiopharmaceutical, 64Cu-cyclam-RAFT-c(-RGDfK-)4 (64Cu-RaftRGD, an αVβ3 integrin [αVβ3] tracer), and 64Cu-diacetyl-bis (N4-methylthiosemicarbazone) (64Cu-ATSM, a supposed tracer for hypoxic metabolism) in a small animal tumor model.
Mice with subcutaneous αVβ3-positive U87MG glioblastoma xenografts were used. The intratumoral distribution of a near-infrared dye, Cy5.5-labeled RAFT-c(-RGDfK-)4 (Cy5.5-RaftRGD), 64Cu-RaftRGD, and 64Cu-ATSM was visualized by fluorescence imaging and autoradiography of the co-injected Cy5.5-RaftRGD with 64Cu-RaftRGD or 64Cu-ATSM at 3 h postinjection. Mice were treated with a single intravenous dose of the vehicle solution (control), 18.5 or 37 MBq of 64Cu-RaftRGD or 64Cu-ATSM, or a combination (18.5 MBq of each agent). The tumor volume, tumor cell proliferation, body weight, survival, and tumor and organ uptake of radiopharmaceuticals were assessed. It was shown that Cy5.5-RaftRGD colocalized with 64Cu-RaftRGD and could be used as a surrogate for the radioactive agent. The intratumoral distribution of Cy5.5-RaftRGD and 64Cu-ATSM was discordant and nearly complementary, indicating a more uniform distribution of radioactivity achievable with the combined use of 64Cu-RaftRGD and 64Cu-ATSM. Neither 64Cu-RaftRGD nor 64Cu-ATSM showed significant effects on tumor growth at 18.5 MBq. The combination of both (18.5 MBq each) showed sustained inhibitory effects against tumor growth and tumor cell proliferation and prolonged the survival of the mice, compared to that by either single agent at 37 MBq. Interestingly, the uptake of the combination by the tumor was higher than that of 64Cu-RaftRGD alone, but lower than that of 64Cu-ATSM alone. The kidneys showed the highest uptake of 64Cu-RaftRGD, whereas the liver exhibited the highest uptake of 64Cu-ATSM. No obvious adverse effects were observed in all treated mice.
The combination of 64Cu-RaftRGD and 64Cu-ATSM achieved an improved antitumor effect owing to the more uniform intratumoral distribution of radioactivity. Thus, combining different radiopharmaceuticals to improve the intratumoral distribution would be a promising concept for more effective and safer TRT.
KeywordsIntratumoral radioactivity distribution Combination targeted radionuclide therapy Tumor co-targeting strategy Therapy response evaluation Radiopharmaceuticals 64Cu-labeled multimeric cRGD peptide 64Cu-ATSM αVβ3 integrin Angiogenesis Hypoxia
Percentage of injected radioactivity per gram of tissue
Hematoxylin and eosin
Positron emission tomography
Targeted radionuclide therapy
Targeted radionuclide therapy (TRT) is a promising cancer treatment modality, which is implemented by systemic administration of a specific targeting vehicle, such as an antibody, peptide, and small organic molecule, that is radiolabeled with an α- or β−-particle emitter or Auger electron emitter to deliver a cytocidal radiation dose to both the primary and metastatic lesions . However, the intrinsically heterogeneous biological properties of tumors would lead to heterogeneous intratumoral distribution of the TRT agent. Thus, delivering sufficient radiation throughout the entire tumor to achieve a greater treatment efficacy remains challenging. If we could find a therapeutic option that could achieve homogeneous intratumoral distribution of radionuclides, this might improve the therapeutic efficacy. There are many types of TRT agents that can target various biological properties; combining different agents may achieve a homogeneous distribution of radionuclides, resulting in delivery of sufficient radiation throughout the entire tumor.
64Cu is a promising theranostic radionuclide owing to its suitable half-life (12.7 h) with multiple decay modes, including β+ (18%), which is used for positron emission tomography (PET) imaging, β− (39%, 0.95−1.4 mm tissue range) , and Auger electron (43%, ~ 126 nm tissue range) , which are used for therapeutic radiation. We previously developed a tetrameric cyclic Arg-Gly-Asp (cRGD) peptide-based radiopharmaceutical, 64Cu-cyclam-RAFT-c(-RGDfK-)4 (64Cu-RaftRGD)  for targeting αVβ3 integrin (αVβ3), a transmembrane glycoprotein receptor that is highly expressed on both the angiogenic endothelial and tumor cells and plays important roles in tumor growth, invasion, metastasis, and angiogenesis [5, 6]. Our previous studies showed that 64Cu-RaftRGD PET could detect αVβ3-positive tumors and angiogenesis, as well as monitor the antiangiogenic effects [4, 7, 8]; in addition, we showed the therapeutic efficacy of 64Cu-RaftRGD as an αVβ3-targeted radionuclide therapy . Moreover, we focused on another Cu(II)-based theranostic agent, 64Cu-diacetyl-bis (N4-methylthiosemicarbazone) (64Cu-ATSM), which supposedly accumulates in regions of hypoxic metabolism within the tumor [10, 11]. As a neutral lipophilic small molecule with high membrane permeability and low redox potential, Cu-ATSM was designed to release Cu(I) ions within cells under over-reduced conditions (such as hypoxia) but not under normal conditions, leading to selective intracellular trapping of Cu(I) ions . Clinical PET studies showed the association between the tumor uptake of 60//62/64Cu-ATSM and therapeutic resistance, metastatic potential, and poor prognosis [10, 11]; in addition, preclinical studies using animal tumor models showed the antineoplastic activity of 64Cu-ATSM [13, 14, 15].
Considering that 64Cu-RaftRGD and 64Cu-ATSM most likely accumulate in different regions of the tumor, with the former accumulating in regions of angiogenesis (toward a normoxic state), and the latter accumulating in regions of hypoxic metabolism, and that angiogenesis and hypoxic metabolism may co-exist since both are common biological features of solid malignancy [16, 17], we hypothesized that a more uniform radioactivity distribution could be achieved via co-injection of the two agents into the tumor. Herein, we aimed to address a new concept for TRT to improve the intratumoral radioactivity distribution using two different radiopharmaceuticals, by studying the combined use of 64Cu-RaftRGD and 64Cu-ATSM in an αVβ3-positive tumor mouse model.
Radiopharmaceuticals and peptide probe
64Cu was produced at the National Institute of Radiological Sciences facility (Chiba, Japan) with a molar activity of 1.77 ± 1.17 TBq/μmol and a radionuclidic purity of > 99.9% . 64Cu labeling of cyclam-RAFT-c(-RGDfK-)4 and H2ATSM was performed, as previously described [4, 19] (brief description in Additional file 1). 64Cu-RaftRGD was produced with a molar activity of 148 MBq/nmol and radiochemical purity of > 99.5%. The radiochemical purity of 64Cu-ATSM was > 95%. A near-infrared dye, Cy5.5-conjugated RAFT-c(-RGDfK-)4 (Cy5.5-RaftRGD) was also prepared, as previously described .
The two radioagents were mixed at a radioactivity ratio of 1:1, kept at room temperature, and analyzed using a reversed-phase high-performance liquid chromatography system, as described in Additional file 1 for determination of the co-administration mode (simultaneously or sequentially).
As previously reported [9, 21], for in vivo use, 64Cu-RaftRGD was formulated in a vehicle solution of normal saline containing 1% Tween 80 to improve the peptide solubility, Gelofusine® (80 mg/kg, Braun Medical, Oss, Netherlands), and l-lysine (400 mg/kg, Sigma-Aldrich, Buchs, Switzerland) to reduce the renal retention of the tracer (by > 50%). 64Cu-ATSM was diluted with normal saline only .
Animals and tumor model
Animal procedures were approved by the Animal Ethics Committee of the National Institutes for Quantum and Radiological Science and Technology (Chiba, Japan) and were conducted in accordance with the institutional guidelines. The αVβ3-positive tumor model was induced by subcutaneous inoculation of 5 × 106 U87MG human glioblastoma cells (American type culture collection) into the right flank of female BALB/cAJcl-nu/nu mice (5 weeks old, CLEA Japan, Inc.), as previously reported . The size of the tumor mass was measured with a digital caliper, tumor volume (mm3) was determined using the following formula: tumor volume = 0.5 × length × width2, and the tumor growth ratio was calculated by dividing the obtained value by the initial tumor volume measured immediately before treatment (day 0). Unless otherwise stated, the mice that had a median body weight value of 21 g (range, 19−23 g) and developed tumors of approximately 100 mm3 at 14 days post-tumor cell inoculation were used for subsequent studies. The humane survival endpoint of mice was determined when the tumor size exceeded 1500 mm3 . It should be emphasized that all animal experiments performed in the present study including those presented in additional files were designed to use the same strain, sex, and age of mice obtained from the same company as abovementioned, and the tumor-bearing mice were randomly divided into groups for subsequent studies.
Intratumoral distribution of radiopharmaceuticals
Mice were pretreated (via the tail vein) with 37 MBq (0.25 nmol) of 64Cu-RaftRGD, and 10 min later, 20 nmol of Cy5.5-RaftRGD was injected (to avoid competitive blocking). Another group of mice bearing tumors of various sizes were co-injected with 37 MBq of 64Cu-ATSM and 20 nmol of Cy5.5-RaftRGD. At 3 h postinjection (p.i.), the mice were euthanized and tumors were excised. Autoradiography, fluorescence microscopic examination, and hematoxylin and eosin (HE) staining were sequentially performed using the same tumor frozen sections. In addition, autoradiography, immunofluorescence staining of the microvascular marker, CD31, and HE staining were performed using adjacent sections. See Additional file 1 for the detailed procedures. The staining for αVβ3 and a hypoxia marker was omitted because previous studies have shown the colocalization of αVβ3 staining and 64Cu-RaftRGD distribution in the tumor model used here  and the correlation of 64Cu-ATSM uptake with the expression of hypoxia-inducible factor-1α (a protein of cellular response to hypoxia) in human tumors  and xenografts .
Mice (n = 5−6/group) were treated with a single administration of the vehicle solution (control), 18.5 or 37 MBq of 64Cu-RaftRGD (0.125 and 0.25 nmol for 18.5 and 37 MBq, respectively) or 64Cu-ATSM (2.5 and 5 nmol for 18.5 and 37 MBq, respectively), or a combination of both at a dose of 18.5 MBq for each agent (radioactivity ratio = 1:1, for the simplicity of study design). Then, the body weight of mice and tumor size were measured every 2 days, and the survival endpoint days (as defined above) were recorded.
In a separate experiment, mice (n = 4/group) were treated with 111 or 148 MBq of the combination to see if the use of higher doses (enhancing the radiation energy) can cause more effective antitumor effect (inducing tumor shrinkage compared to the tumor size before treatment). Hematology and hepatorenal function tests were carried out for toxicity evaluation, as described in Additional file 1.
Mice (n = 5/group) were treated with 0.74 MBq of 64Cu-RaftRGD (0.25 nmol), 64Cu-ATSM (5 nmol), or the combination (0.125 nmol of 64Cu-RaftRGD plus 2.5 nmol of 64Cu-ATSM). The “nmol” amounts indicated above were the values adjusted to those used for the 37 MBq therapy protocols. At 3 and 24 h p.i., the mice were euthanized, and the blood, tumor, muscles, liver, and kidneys were collected and measured using a gamma counter with decay correction. Tumor and organ uptake was expressed as the percentage of injected radioactivity per gram of tissue (%IA/g) normalized to 20 g body weight by multiplying with the ratio of body weight to 20 g.
Histological evaluation of tumor proliferation
Mice (n = 3−4/group at each time point) were injected with the vehicle solution or 37 MBq of 64Cu-RaftRGD, 64Cu-ATSM, or the combination. At 1 and 3 days p.i., the mice were euthanized and tumors were excised for immunohistochemical staining of the proliferation marker, Ki67. As previously described , the proliferation index was defined as the percentage of Ki67-positive cells, which was quantified by counting five areas (× 40 objective lens) of highest Ki67 staining intensity using the WinROOF 2015 image analysis software (version: 3.10.0, Mitani Corporation, Fukui, Japan). See Additional file 1 for the detailed procedures.
Quantitative data were presented as the means ± standard deviations. Unpaired t test and one-way analysis of variance followed by Dunnett’s test were used for two-group and multiple comparisons, respectively (Kaleida Graph 4.0). Based on the period from day 0 p.i. to the specified survival endpoint day, Kaplan-Meier survival curves were plotted, and the differences in survival were evaluated using the log-rank test (GraphPad Prism 5). P values < 0.05 were considered statistically significant.
Intratumoral distribution of 64Cu-RaftRGD and 64Cu-ATSM
Radiotherapy with 64Cu-RaftRGD, 64Cu-ATSM, and a combination of both
In comparison with the 37 MBq combination (Additional file 6a), increasing the total radioactivity dose of the combination to 111−148 MBq induced significant tumor shrinkage in a dose-dependent manner (Additional file 6b); additionally, it did not cause a marked or long-lasting body weight loss (Additional file 6c). The 148 MBq-treated mice showed a maximal mean body weight reduction of 9.9% (less than 10%) on day 6, compared to day 0 (baseline), and afterwards the body weight gradually recovered beyond the baseline. Measurement of hematology and hepatorenal functions (Additional file 7) did not show a dose-response relationship between the 111 and 148 MBq doses. All the values obtained on day 21 after treatment were not significantly different from those measured in the control mice.
Histological study of tumor proliferation
Biodistribution of 64Cu-RaftRGD, 64Cu-ATSM, and a combination of both
Biodistribution of 64Cu-RaftRGD, 64Cu-ATSM, and a combination of both in U87MG tumor-bearing mice
3 h p.i.
24 h p.i.
(“a” + “b”)/2
(“a” + “b”)/2
0.05 ± 0.01
2.1 ± 0.12
1.15 ± 0.14*a, *b
0.04 ± 0.01
1.68 ± 0.15
0.83 ± 0.06*a, *b
8.06 ± 0.56
19.9 ± 1.41
12.8 ± 0.42*a, *b
5.43 ± 0.27
12.1 ± 0.69
8.48 ± 0.28*a, *b
11.8 ± 0.86
8.38 ± 0.49
9.55 ± 0.79†a, §b
8.42 ± 0.61
6.98 ± 0.34
7.18 ± 0.38‡a,║b
0.4 ± 0.06
0.73 ± 0.06
0.57 ± 0.06†a, ‡b
0.24 ± 0.04
0.63 ± 0.05
0.46 ± 0.06*a, †b
6.69 ± 0.64
15.5 ± 2.03
10.7 ± 1.05‡a, †b
5.18 ± 0.24
9.47 ± 0.78
7.02 ± 1.2‡a, ‡b
To the best of our knowledge, the present proof-of-concept study is the first to report the use of a combination of different radiopharmaceuticals to improve the heterogeneous intratumoral distribution of radionuclides and enhance the therapeutic efficacy. So far, the most commonly used strategy to improve the efficacy of TRT involves its combination with other treatment modalities, such as chemotherapy and external radiotherapy .
It is of great importance to determine a suitable co-administration mode (simultaneous or sequential) for combination of different drugs owing to the possibility of in vitro and in vivo drug-drug interactions, as well as the difference in biopharmacokinetics of the drugs. In the present study, we selected the simultaneous mode based on the results of the chromatographic analysis of 64Cu-RaftRGD and 64Cu-ATSM mixture, which did not indicate any radioagent-radioagent reaction, such as transchelation (owing to the potentially different 64Cu-coordinating ability of the two agents) within 1 h after mixing the two agents. In addition, our previous data did not show different biodistribution patterns of the two radiopharmaceuticals [21, 22]. In addition, the results of the present biodistribution study using the same experimental settings further showed no radioagent-radioagent reactions in vivo, as evidenced by the fact that the measured tumor and organ uptake values for the co-injected 64Cu-RaftRGD and 64Cu-ATSM were approximately equal to the average values of those obtained for the single agents.
Expectedly, the intratumoral distribution study showed spatially discordant but complementary distribution patterns of 64Cu-RaftRGD and 64Cu-ATSM in the αVβ3-positive U87MG tumor xenograft, which resulted in a more uniform radioactivity distribution in the tumors co-injected with the two agents. Immunostaining with CD31 revealed that the regions of high 64Cu-ATSM and low 64Cu-RaftRGD accumulation were enriched with relatively smaller microvessels than the regions of low 64Cu-ATSM and high 64Cu-RaftRGD accumulation. This difference in the microvasculature indicated the complementary relationship between 64Cu-ATSM and 64Cu-RaftRGD accumulation according to the intratumoral vascularization levels, which is in accordance with our proof of concept in the present study.
In the subsequent therapeutic study, despite using the same total radioactivity dose (37 MBq), the combined use of 64Cu-RaftRGD and 64Cu-ATSM (18.5 MBq of each agent) was more efficient in maintaining its inhibitory effects against tumor growth and tumor cell proliferation and prolonging the survival of mice than either single agent. At a dose of 18.5 MBq, neither 64Cu-RaftRGD nor 64Cu-ATSM showed significant effects on tumor growth. Hence, these results might indicate a synergistic antitumor effect between 64Cu-RaftRGD and 64Cu-ATSM. We further examined the total tumor radioactivity uptake of 64Cu-RaftRGD, 64Cu-ATSM, and the combination at the therapeutic dose. Lower tumor uptake values were observed for the more effective combination therapy (10.7 ± 1.05 and 7.02 ± 1.20 %IA/g at 3 and 24 h, respectively), compared to the corresponding values of 64Cu-ATSM alone (15.5 ± 2.03 and 9.47 ± 0.78 %IA/g, respectively), indicating no decisive correlation between the tumor uptake and antitumor effects. Taken together, results of the intratumoral distribution, total tumor uptake levels, and therapeutic effects suggested that the improved antitumor effects of the combination therapy could be largely attributed to the improved intratumoral radioactivity distribution. Regarding the therapeutic efficacy of 64Cu-RaftRGD plus 64Cu-ATSM combination, a dose-response relationship was revealed; a total dose of 37 MBq induced tumor growth delay, whereas higher doses (111 and 148 MBq) caused tumor growth inhibition. It should be mentioned that to simplify the experiment schedules, the mice were treated with only one single-dose injection, but despite this, significant tumor growth delay and promising results of tumor growth inhibition were achieved by the combination with the total doses of 37 MBq and 111−148 MBq, respectively. Actually, in clinical settings, TRT as well as other treatment modalities such as chemotherapy and external radiotherapy are administered or performed multiple times to obtain the most effective therapeutic efficacy. Thus, it may be anticipated that multiple-dose regimen of the combined TRT with 64Cu-RaftRGD and 64Cu-ATSM could greatly improve the treatment effectiveness.
For TRT, the presence of dose-limiting organs usually restricts the administered activity levels of therapeutic radiopharmaceuticals. Radiopharmaceuticals considered for combined use should have different critical organ profiles, which may lower the risk of dose-limiting toxicity and widen the therapeutic window. Previous studies showed that the kidneys constitute the principal dose-limiting organ for 64Cu-RaftRGD therapy , whereas the liver, red marrow, and ovaries are the dose-limiting organs for 64Cu-ATSM therapy . In the present study, the kidneys exhibited the highest tissue uptake of 64Cu-RaftRGD followed by the liver, whereas the liver showed the highest tissue uptake of 64Cu-ATSM followed by the kidneys. Throughout the study (up to 3−4 weeks after treatment), we did not observe any serious side effects even in the mice treated with the highest dose (148 MBq; 74 MBq of 64Cu-RaftRGD plus 74 MBq of 64Cu-ATSM). Our previous study showed that 64Cu-ATSM at 148 MBq caused hematological toxicity, which might be related to the resultant poor survival of the treated mice . Although further long-term evaluation of safety is needed, the combination therapy has the potential to be safer, compared to dose escalation of a single agent.
So far, only few studies have reported the combined use of different radiopharmaceuticals for TRT . Previous studies performed TRT of neuroendocrine tumors via the combined use of high-energy and long-path 90Y-somatostatin analog with its medium energy and short path 177Lu analog for combating tumors of various sizes , or with 131I-metaiodobenzylguanidine (a norepinephrine analog) for increasing the dose delivered to the tumor site . Different from their treatment concepts, the present study suggested the combination of different radiopharmaceuticals (may not be limited to the currently used agents) to improve the intratumoral radioactivity distribution and thus enhance the therapeutic efficacy. Despite different concepts, all combined TRT treatments might have a common advantage: allowance of personalized and optimized dose calculation for each radiopharmaceutical based on the dosimetric data and spatial distribution in the tumor obtained from separate PET scans because many TRT agents or their surrogates are also PET tracers.
The performance of the proposed newly designed treatment concept should be evaluated in further studies using various tumor models and, if possible, different combinations of radioagents with optimized doses for each single agent to maximize the therapeutic efficacy. Another limitation of the present study is the lack of information on the radiation absorbed dose distribution in tumors, which might be more directly related to the tumor response. However, to obtain these data, a sophisticated three-dimensional dose calculation technique would be needed .
In conclusion, the combined use of two different radioagents, 64Cu-RaftRGD and 64Cu-ATSM with discordant intratumoral distribution patterns resulted in a more uniform intratumoral radioactivity distribution, resulting in an improved antitumor effect. Combining different radiopharmaceuticals to improve the intratumoral distribution would be a promising concept for more effective and safer TRT.
We would like to thank the Department of Radiopharmaceuticals Development, National Institute of Radiological Sciences (Chiba, Japan), for supplying the 64Cu used in this study and the Cyclotron Operation Section for cyclotron operation.
This research did not receive any specific grant from any funding agency in the public, commercial, or non-profit sector.
Availability of data and materials
Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.
ZHJ and ABT conceived and designed the present study. ZHJ, ABT, YF, and DB developed methodology. ZHJ, MD, AS, KN, and MRZ acquired the data. ZHJ, ABT, MD, and YY analyzed and interpreted the data. MD, YY, KN, PD, and DB provided the technical and material support. AS and MRZ provided administrative and material support. ZHJ created the figures and tables, drafted and revised the manuscript. ABT, YF, PD, DB, and TH reviewed and revised the manuscript. ABT and TH supervised the overall project. All authors read and approved the final manuscript.
The animal experiments in this study were performed in accordance with institutional and national guidelines regarding animal care and approved by the Animal Ethics Committee of the National Institutes for Quantum and Radiological Science and Technology (Chiba, Japan). This article does not describe any studies with human participants performed by any of the authors.
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 4.Jin ZH, Furukawa T, Galibert M, et al. Noninvasive visualization and quantification of tumor αVβ3 integrin expression using a novel positron emission tomography probe, 64Cu-cyclam-RAFT-c(-RGDfK-)4. Nucl Med Biol. 2011;38(4):529−40.Google Scholar
- 8.Aung W, Jin ZH, Furukawa T, et al. Micro-positron emission tomography/contrast-enhanced computed tomography imaging of orthotopic pancreatic tumor-bearing mice using the αVβ3 integrin tracer 64Cu-labeled cyclam-RAFT-c(-RGDfK-)4. Mol Imaging. 2013;12(6):376−87.Google Scholar
- 9.Jin ZH, Furukawa T, Degardin M, et al. αVβ3 Integrin-targeted radionuclide therapy with 64Cu-cyclam-RAFT-c(-RGDfK-)4. Mol Cancer Ther. 2016;15(9):2076−85.Google Scholar
- 10.Vavere AL, Lewis JS. Cu-ATSM: a radiopharmaceutical for the PET imaging of hypoxia. Dalton Trans. 2007;43:4893−902.Google Scholar
- 20.Jin ZH, Josserand V, Foillard S, et al. In vivo optical imaging of integrin αV-β3 in mice using multivalent or monovalent cRGD targeting vectors. Mol Cancer. 2007;6:41.Google Scholar
- 21.Jin ZH, Furukawa T, Sogawa C, et al. PET imaging and biodistribution analysis of the effects of succinylated gelatin combined with L-lysine on renal uptake and retention of 64Cu-cyclam-RAFT-c(-RGDfK-)4 in vivo. Eur J Pharm Biopharm. 2014;86(3):478−86.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.